Skip to main content

Full text of "The applications of electrolysis in chemical industry"

See other formats

Monographs on Industrial Chemistry 




xfc, ~ 


Edited by Sir EDWARD THORPE, C.B., LL.D., F.R.S. 

Emeritus Professor of General Chemistry in the Imperial College of Science and Technology, 
South Kensington ; and formerly Principal of the Government Laboratory, London. 


TOURING the last four or five decades the Appli- 
^-^^ cations of Chemistry have experienced an extra- 
ordinary development, and there is scarcely an industry 
that has not benefited, directly or indirectly, from this 
expansion. Indeed, the Science trenches in greater 
or less degree upon all departments of human activity. 
Practically every division of Natural Science has now 
been linked up with it in the common service of man- 
kind. So ceaseless and rapid is this expansion that 
the recondite knowledge of one generation becomes a 
part of the technology of the next. Thus the conceptions 
of chemical dynamics of one decade become translated 
into the current practice of its successor ; the doctrines 
concerning chemical structure and constitution of one 
period form the basis of large-scale synthetical processes 
of another ; an obscure phenomenon like Catalysis is 
found to be capable of widespread application in 
manufacturing operations of the most diverse character. 
This series of Monographs will afford illustrations of 
these and similar facts, and incidentally indicate their 
bearing on the trend of industrial chemistry in the near 
future. They will serve to show how fundamental and 
essential is the relation of principle to practice They 

will afford examples of the application of recent know- 
ledge to modern manufacturing procedure. As regards 
their scope, it should be stated the books are not intended 
to cover the whole ground of the technology of the matters 
to which they relate. They are not concerned with the 
technical minutia of manufacture except in so far as these 
may be necessary to elucidate some point of principle. In 
some cases, where the subjects touch the actual frontiers of 
progress, knowledge is so very recent and its application 
so very tentative that both are almost certain to ex- 
perience profound modification sooner or later. This, 
of course, is inevitable. But even so such books have 
more than an ephemeral interest. They are valuable as 
indicating new and only partially occupied territory ; and 
as illustrating the vast potentiality of fruitful conceptions 
and the worth of general principles which have shown 
themselves capable of useful service. 

Organic Compounds of Arsenic and Antimony. By G. T. 

MORGAN, F.R.S., F.I.C., M.R.I.A., D.Sc., A.R.C.Sc., Professor of 

Applied Chemistry, City and Guilds Technical College, Finsbury, 

London. i6s. net. 

Edible Oils and Fats. By C. A. MITCHELL, F.I.C. 6s. 6d. net. 
Coal and its Scientific Uses. By W. A. BONE, D.Sc., F.R.S., 

Imperial College of Science and Technology, South Kensington. 

2is. net. 
The Zinc Industry. By ERNEST A. SMITH, The Assay Office, 

Sheffield. los. 6d. net. 
Colour in Relation to Chemical Constitution. By E. R. 

WATSON, M.A., D.Sc., Professor of Chemistry, Dacca College, 

Bengal. 1 2 s. 6d. net. 
The Applications of Electrolysis in Chemical Industry. By 

ARTHUR J. HALE, B.Sc., F.I.C., Finsbury Technical College, 

The Natural Organic Colouring Matters. By A. G. PERKIN, 

F.R.S., The Dyeing Department, The University, Leeds; and 

A. E. EVEREST, D.Sc., PH.D., Technical College, Huddersfield. 
Catalysis in Industrial Chemistry. By G. G. HENDERSON, M.A., 

D.Sc., LL.D., F.R.S., The Royal Technical College, Glasgow. 
The following Volumes are in preparation: 
Liquid Fuel for Internal Combustion Engines. By Sir BOVER- 

TON REDWOOD, Bart., D.Sc., F.R.S.E., and J. S. S. BRAME, Royal 

Naval College, Greenwich. 

Synthetic Colouring: Matters: Sulphur Dyes. By G. T. 

MORGAN, D.Sc., A.R.C.S., F.R.S., Finsbury Technical College, 

Synthetic Colouring Matters: Vat Colours. By JOCELYN F. 
THORPE, C.B.E., D.Sc., F.R.S., Imperial College of Science arid 
Technology, South Kensington. 

Naphthalene. By W. P. WYNNE, D.Sc., F.R.S., The University, 

Synthetic Colouring Matters : Azo-Dyes. By FRANCIS W. KAY, 

D.Sc., The University, 'Liverpool. 

Utilisation of Atmospheric Nitrogen : Synthetical Production 

of Ammonia and Nitric Acid. By A. W. CROSSLEY, C.M.G., 
D.Sc., F.R.S., F.I.C., King's College, Strand. 


The Principles and Practice of Gas -purification. By EDWARD 
V. EVANS, F.I.C., Chief Chemist, South Metropolitan Gas Company. 

Refractories. By J. W. MELLOR, D.Sc. 

Ozone and Hydrogen Peroxide: their Properties, Technical 
Production and Applications. By H. VINCENT A. BRISCOE, 
D.Sc., A.R.C.S., Imperial College of Science and Technology, 
South Kensington. 

The Nickel Industry. By WILLIAM G. WAGNER. 
Cellulose- Silk. By C. F. CROSS, B.Sc., F.R.S., F.I.C. 

The Electric Arc in Chemical Industry. By J. N. PRING, D.Sc., 

The University, Manchester. 
By- Product Coking Practice. By ERNEST BURY, M.Sc. 

Organic Synthetic Reactions : their Application to Chemical 
Industry. By JULIUS B. COHEN, B.Sc., Ph.D., F.R.S. 

Synthetic Colouring Matters: Triphenylmethane Dyes. By 

R. ROBINSON, D.Sc., Professor of Organic Chemistry in the 
University of Liverpool. 

Synthetic Colouring Matters: Anthracene and Allied Dye- 
stuffs. By F. W. ATACK, M.Sc. Tech., B.Sc. (Lond.), F.I.C. 
of the Municipal School of Technology, Manchester. 

Synthetic Colouring Matters: Acridine and Xanthene Dye- 
stuffs. By JOHN T. HEWITT, M.A., D.Sc., F.R.S., University of 
London (East London College). 

Synthetic Colouring Matters: Azine and Oxazine Dye-stuffs. 

By JOHN T. HEWITT, M.A., D.Sc., F.R.S., University of London 
(East London College). 

Synthetic Drugs: Local Anaesthetics. By W. H. HURTLEY, 
D.Sc., St. Bartholomew's Hospital ; and M. A. WHITELEY, D.Sc., 
Imperial College of Science and Technology, South Kensington. 








Demonstrator and Lecturer in Chemistry 
The City and Guilds of London Technical College^ Finsbury 






THE scope and purpose of this volume are sufficiently 
indicated by its title. Electrolysis now plays an important 
part in the processes of Chemical Industry, and the value 
of Electro-Chemistry is generally recognised. 

It is hoped that this work will prove useful to all those 
who are associated in any way with Chemical Science, and 
that it may stimulate interest in a rapidly growing branch 
of Chemistry which is worthy of more serious attention. 

An account of the general principles of electrolysis and 
an explanation of the terms relating thereto have been given 
in the Introduction, and, at the suggestion of the Editor, 
a chapter has been included on methods of generating 

These two sections should prove serviceable to the 
student and the general reader, and the numerous references 
to original papers will give those who may desire it an 
introduction to the literature of the subject. 

The Author wishes to record his indebtedness to Mr. 
L. W. Phillips, A.M.I.E.E., for valuable assistance in connec- 
tion with the chapter on methods of generating current. 

LONDON, July igi8. 




Electrolysis The ionic theory Faraday's laws Osmotic pres- 
sure Electrical units Decomposition voltage Electrical 
osmosis Cataphoresis Colloids in electrolysis Current effi- 
ciency Energy efficiency Electrolysis bath Electrodes 
Diaphragms Molten electrolytes 


Primary Cells : Simple voltaic cell Daniell's cell Leclanche cell 
Lelande cell Fuel cells. Secondary cells : Reactions of lead 
accumulator The thermopile Dynamo-electric machines 
Alternating and direct current The rotary converter Motor 
generator Transformers Measurement of current Ammeters 
and voltmeters. Power and electro-chemical industry : Costs 
of power from various sources Power prospects in the United 


Copper: Multiple and series systems. Lead: Fluosilicate pro- 
cess Perchlorate process. Tin. Iron. Cadmium. Bullion 
refining (Silver and Gold] : Process of Moebius Philadelphia 
Mint Raritan copper works process Balbach-Thum process 
Wohlwill process for gold. 


Aluminium. Copper: Marchese process Process of Siemens 
and Halske Hoepfner process. Zinc : Electrolysis of aqueous 
solutions Electrolysis of fused zinc chloride. Lead: Salom's 
process. Nickel: Hoepfner process Process of Savelsburg 
and Wannschaff Browne process. Sodium: Castner process 
Darling process Contact Electrode process Ashcroft pro- 
cess. Magnesium. Calcium* Lithium. Antimony. Bismuth. 






Quantitative relations Historical Schmidt's process Schoop's 
process Process of Garuti Schuckert process Cell of the 
International Oxygen Company Modern filter-press cells 
Electrolytic production of ozone. 



SODA ........... 90 

General principles Diaphragm cells: Griesheim cell Har- 
greaves-Bird process Outhenin-Chalandre cell Townsend 
cell Finlay cell MacDonald cell Le Seur cell Billiter- 
Siemens cell. Mercury cells: Castner-Kellner cell Solvay 
cell Whiting cell Rhodin cell Wilderman cell. The Bell 
Process: Aussig bell process Billiter-Leykam cell. Elec- 
trolysis of fused salt: The Acker process. Present position 
and future of Electrolytic Alkali. 



General Principles Kellner cell for hypochlorite Schuckert cell 
Haas-Oettel cell Schoop cell for hypochlorite. Chlorate 
production : Gibb's process Process of Lederlin and Corbin 
Perchlorates Bromine and bromates Iodine. 


White lead Lead peroxide Chrome yellow Cuprous oxide 
Percarbonates Hydroxylamine Hydrosulphites Persulph- 
uric acid and hydrogen peroxide Potassium permanganate 
Potassium ferricyanide Nitric acid Fluorine. 


lodoform Anthraquinone Vanillin I sopropyl alcohol Chloral 
Saccharine Reduction of nitro-compounds Electrolytic 
oxidation of organic compounds Coal tar dyes, 


NAME INDEX . . . 147 


Amer. Chem. Journ. 
Ber. . . . . 

Bull, de F Assoc. Ing. 

Chem. Zeit. 
Chem. Trade Journ. . 
Compt. rend. . 


Elect, and Met. Ind. 
Electrochem. Ind. 
Electrochem. Review 
Eng. and Mining Journ. . 
Eng. Pat. 

Fr. Pat 

Int. Cong. App. Chem. 
Journ. hid. and Eng. Chem. 

Journ. pr. Chem. 
Journ. phys. Chem. . 
Journ. Soc. Chem. Ind. 
Journ. Soc. Dyers and 


Mel. and Chem. Eng. 
Monit. Scient. . 
Trans. Amer. Electrochem. 

Trans. Faraday Soc. 
U.S. Pat. 

Zeitsch. angew. Chem. 
Zeitsch. anorg. Chem. 
Zeitsch. Elektrochem. 
Zeitsch. phys. Chem. 


American Chemical Journal. 
Berichte der Deutschen chemischen Gesell- 

Bulletins de 1'Association des Ingenieurs 

Chemiker Zeitung. 
Chemical Trade Journal. 
Comptes rendus hebdomadaires des Stances 

de 1' Academic des Sciences. 
Deutsches Reichspatent. 
Electrochemical and Metallurgical Industry. 
Electrochemical Industry. 
Electrochemical Review. 
Engineering and Mining Journal. 
English Patent, 
French Patent. 

International Congress of Applied Chemistry. 
Journal of Industrial and Engineering 


Journal fur praktische Chemie. 
Journal of Physical Chemistry. 
Journal of the Society of Chemical Industry. 
Journal of the Society of Dyers and Colorists. 

Metallurgical and Chemical Engineering. 
Moniteur Scientifique de Quesneville. 
Transactions of the American Electro- 
chemical Society. 

Transactions of the Faraday Society. 
United States Patent. 
Zeitschrift fiir angewandte Chemie. 
Zeitschrift fiir anorganische Chemie. 
Zeitschrift fiir Elektrochemie. 
Zeitschrift fiir physikalische Chemie. 




ELECTROLYSIS is the term given to the process by which 
a compound is decomposed, when in solution, by the passage 
through it of an electric current. 

The compound is dissolved for. this purpose in some liquid 
medium, usually water, but it is sometimes possible to use the 
compound alone, in a molten state. 

In the early years of the nineteenth century, Humphry 
Davy and Michael Faraday investigated the subject of 
electrolysis, and established many useful facts. By 1880, it 
was realised that the electrolytic decomposition of substances 
was of industrial importance, and processes were soon devised 
and patented, in Europe and America, which involved the 
application of electrolysis in chemical industry. 

It is essential to discuss, briefly, the general facts and 
principles of electrolysis, before passing to a detailed study 
of the various industrial processes. 

When an electric current traverses a solution (a liquid 
conductor), its passage is accompanied by the decomposition 
of the substance, whereas an ordinary metal conductor is not 
decomposed by the passage of electricity through it. 

The liquid is termed the electrolyte, and the terminals 
immersed in it, by which the current enters and leaves, are 
called the electrodes. 

That electrode which is connected to the positive pole of 

the battery or generator is the anode, and the other, which is 


connected : to the" negative .pole of the current source, is known 
as the cathode ; the current is regarded as entering the 
electrolyte at the anode and leaving it at the cathode. 

By electrolysing an aqueous solution of copper sulphate 
between platinum electrodes, metallic copper is deposited at 
the cathode whilst oxygen is evolved at the anode ; similarly, 
the electrolysis of acidified water yields hydrogen at the 
cathode and oxygen at the anode. Sodium chloride in aque- 
ous solution yields hydrogen at the cathode and chlorine at 
the anode, but if molten sodium chloride be used, metallic 
sodium will be deposited on the cathode. 

Evidently, secondary reactions take place under certain 
conditions, because, instead of the expected sodium, hydrogen 
gas in equivalent amount is produced at the cathode, when 
water is present, during the electrolysis of sodium chloride. 
This is due to the chemical reaction of the liberated sodium 
with the water thus 

2Na + 2H 2 O = 2NaOH + H 2 . 

There is also a secondary reaction at the anode in the 
case of aqueous copper sulphate, for in place of the complex 
, oxygen is obtained, owing to the reaction between 
and water 

Hydrogen and the metals are liberated at the cathode, 
whilst oxygen and the non-metallic elements are liberated 
at the anode. 

Probably the current is conveyed or conducted through 
the solution by the positive and negative components into 
which the compound is resolved when in the dissolved state. 

This idea is due to Faraday (1834), and he termed these 
carriers of electricity ions; those discharged at the cathode 
are cathions, and those discharged at the anode are anions. 
The cathions carry positive charges of electricity, the anions 
carry negative charges, and when the ions reach their respec- 
tive electrodes their charges are given up or neutralised, and 

the substance is liberated or discharged. The ions are denoted 

+ + ++ 
by symbols in this manner : H', Na', Cu'; or H, Na, Cu ; 


which symbols signify that the ions of hydrogen and sodium 
carry one positive charge, and the copper ion two positive 
charges. Anions are denoted thus : Cl' or Cl ; the chlorine 
ion carries one negative charge. 

There is evidence that the substance which conducts the 
current, when dissolved in water, is not "split up" or dis- 
sociated by the energy of the current, but rather, in the solu- 
tion it exists, to a considerable extent, in the dissociated or 
ionised condition. Since Joule's Law (H = l*rt) holds for 
liquid conductors, the energy of the current produces heat 
in the conductor, and is not used in supplying the disruptive 
force necessary to decompose the substance. If the ions are 
already present in the solution owing to the dissociating or 
ionising influence of the solvent, then, as soon as a difference 
of potential is set up at the electrodes, the charged ions will 
travel towards their respective poles. The positively charged 
ions, for instance, travel to the cathode, which is negative, 
where they lose their charges and are liberated. 

Faraday showed that the quantity of electrolyte decom- 
posed is proportional to the quantity of electricity which 
passes through it, and he also proved that when a current 
passes through several electrolytes in series, the different 
weights of the elements liberated at the electrodes are in 
the same ratio as the chemical equivalents. Hence, the same 
amount of electricity is required to discharge one chemical 
equivalent of any element, and is approximately 96,500 

Evidence of the existence of ions, in aqueous solutions, 
is obtained by measurements of osmotic pressure, and similar 
evidence results from a study of the lowering of freezing 
point or elevation of boiling point of aqueous solutions. It 
is conceivable that in dilute solutions, dissolved substances 
distribute themselves throughout the solvent as the mole- 
cules of a gas distribute themselves by diffusion and fill 
any space. 

Diffusion takes place in liquids, because, if two solutions 
are placed carefully in contact one above the other, the two 
layers after a time disappear and the composition of the 


liquid is the same throughout. The molecules exert a pres- 
sure (osmotic pressure) analogous to gas pressure, and this 
pressure can be measured by means of apparatus arranged 
as shown in Fig. I. A porous pot is fitted with a manometer 
tube which is fixed in a rubber cork, fitting tightly into the 
porous pot. The pot is previously provided with a semi- 
permeable membrane in its wall, by filling it with potassium 
ferrocyanide, and immersing it in a solution of copper sul- 
phate. Where the solutions meet in the interstices of the 
pot, a deposit of copper ferrocyanide is formed which allows 
free passage for water, but not for sugar molecules dissolved 
in the water. 

if the pot, thus prepared, be filled 
with concentrated sugar solution, the 
manometer tube fixed and sealed off 
at B, and the pot then immersed in a 
beaker of water, after standing some 
time a difference of level in the 
manometer liquid will indicate that 
pressure is exerted inside the pot. 

The pressure is due to the effort of 
the sugar molecules to get into the 
water, and so become diluted, just as 
a gas tends to pass to a vacuum and so reduce its pressure. 

The sugar molecules are unable to pass out, therefore dilu- 
tion is brought about by water passing in, and this continues 
until the pressure above the liquid in the pot is equal to 
the pressure in the water outside. The manometer gives a 
measure of the osmotic pressure of the solution in its diluted 

J. H. van't Hoff (1887) showed that in dilute solutions, 
Boyle's Law is true for osmotic pressure as it is for gas 
pressures. He used Pfeffer's measurements of osmotic 
pressure to show that the pressure is proportional to 

The following are some measurements made by H. N. 
Morse (1907) l 

1 Amer. Chem.Journ^ 1907, 37, 324. 

FIG. i. 


Cone, of Molecules 








per litre 

Osmotic Pressure in 









Equivalent Gas Pres- 









The figures for the two pressures are parallel. 

It has similarly been shown, that in dilute solutions 
Charles's Law is valid and that osmotic pressure is directly 
proportional to the absolute temperature. 

Avogadro's hypothesis is therefore applicable to dilute 
solutions, and hence solutions which have the same molecular 
concentration of solute will have the same osmotic pressure. 

Now, passing to consider electrolytes, it is found that 
many substances give abnormally high pressure values. 

Arrhenius determined the number of molecules present in 
dilute solutions of the following substances, when quantities 
proportional to the molecular weights were dissolved in water 


Meihyl alcohol . 0*94 
Cane sugar . . I -o 
Ethyl acetate . 0*96 


MgS0 4 

SnCl 2 



The conclusion usually accepted is, that the inorganic 
substances (salts) are partly ionised or undergo electrolytic 
dissociation when dissolved. 

As already mentioned, the cathions are discharged at the 
cathode, giving a -f charge to the electrode. The opposite 
exchange is going on at the anode, and the source of current 
is necessary to maintain the electrodes at a difference of 
potential so that anions and cathions may be continuously 

Energy is consumed in this work, and not in separating 
the molecules of solute into two or more parts ; possibly the 
heat of hydration of the ions furnishes the energy necessary 
for ionisation (see Decomposition Voltage). 



Electrical energy comprises two factors, quantity and 
intensity, as in water-flow, the quantity is determined by the 

In electrical measurements pressure is often termed 
potential or electromotive force (E.M.F.). 

The quantity of electricity which passes a certain point in 
a conductor in unit time is the current, and this is analogous 
to rate of flow of water. The rate is determined by the 
difference of potential at the ends of the conductor, and by 
the resistance of the system ; the relation between these 
quantities is known as Ohm's Law, expressed thus : I = E/R. 

The unit of quantity is the coulomb, that which deposits 
ooi 1 1 8 gm. of silver, and a current carrying one coulomb per 
second is one ampere. In other words, the ampere is that 
current which passed through a specified solution of silver 
nitrate deposits '001118 gm. of silver per second. 

The solution of silver nitrate is specified thus : It contains 
10-20 gms. of silver nitrate in 100 gms. of distilled water; 
not more than 30 per cent, of the silver may be deposited 
when the test is running, and the cathode current density 
must not be greater than *O2 amp. per cm 2 . 

The unit of resistance is the ohm and is equal to the re- 
sistance of a column of mercury of I mm 2 , cross section, having 
a length of 106*30 cms. and a weight of 14*452 gms. at o C. 

The unit of E.M.F. is the volt, the pressure necessary to 
send a current of one ampere through a resistance of one ohm. 

The Weston cadmium cell is the usually accepted standard 
of electromotive force; at 20 C. its potential is 1*0184 volts, 
and its temperature coefficient is very small. 

One faraday of electricity is a quantity equal to 96,500 
coulombs, and the passage of this amount of electricity through 
a cell deposits or liberates one gram-equivalent of an element. 

The unit of power is the watt; it is the rate at which 
energy is expended by an unvarying current of one ampere 
flowing under a pressure of one volt. 

The corresponding unit of energy is the watt-second or 


joule, that is, the energy expended in one second, when one 
ampere flows under a pressure of one volt. 

For industrial purposes these units are too small and the 
following multiples are in use 

The kilowatt = 1000 watts ; the horse-power = 746 watts. 
Corresponding energy units are the kilowatt-hour (K.W.H.) 
and the horse-power-hour (H.P.H.). The kilowatt- hour is 
also the Board of Trade unit. 

Relations between heat energy and electrical energy which 
are constantly made use of are, one watt-second or a joule =* 
239 gram-calorie, therefore one gram-calorie = 4*189 joules. 

It is possible, by means of these relations, to calculate from 
the heat of formation of a compound the voltage necessary to 
decompose it. For example, the heat of formation of one 
gram-molecule of sodium chloride is 97,690 calories, and this 
energy is equal to 97,690 X 4^19 joules or volt-coulombs. Now 
in order to decompose 585 gms. of sodium chloride the 
quantity of electricity required is 96,500 coulombs, and 

therefore the voltage will be 97)69 X 4>I9 = 4-22 volts. 


At its fusion temperature the voltage required will not be 
so great, because the heat of formation is less, by the quantity 
needed to raise 58 - 5 gms. of sodium chloride from 15 to 
772 C., the specific heat of the chloride being "214, that is 

Heat required = 58-5 x 757 X '214 = 9480 calories 
.'. heat of formation = 97,690 9480 = 88,210 calories. 

TT -11 1 88,2IO X 4*19 

Hence the voltage required will be = 3-81. 


These voltages, calculated from heats of formation, are 
minimum values not realised in practice, because owing to 
various conductivity losses they are always exceeded. 


For the decomposition of every electrolyte a definite 
E.M.F. must be maintained between the electrodes if the 
anions and cathions are to be continuously separated. When 
for example, dilute sulphuric acid is electrolysed between 


platinum electrodes, a back E.M.F. is developed by the layers 
of gas which collect on anode and cathode, and this will 
ultimately reach 17 volts, therefore a voltage greater than 
this must be used if electrolysis is to continue. 

The decomposition voltages for a few electrolytes are : 
ZnS0 4 = 2- 3 5; NiSO 4 = 2-09 ; Pb(NO 3 ) 2 = 1-52 ; AgNO 3 
= o 70. Dilute acids and bases have a decomposition voltage 
of approximately 170 volts, and the products of electrolysis 
are oxygen and hydrogen. 

The greater part of the decomposition voltage needed for 
continuous electrolysis is utilised in overcoming polarisation 
effects due to the deposition of gas or metal on the electrodes. 
During the electrolysis of a dilute solution of acid or alkali 
between platinum electrodes, hydrogen gas forms a layer on 
the cathode, and oxygen a layer on the anode, with the result 
that instead of two platinum plates there are two of different 
materials, one of oxygen, the other of hydrogen, in dilute acid 
or alkali, and the arrangement acts like an accumulator, until 
the gases have passed back again into the solution, giving a 
current in a direction opposite to that first passed through. 

When copper sulphate is electrolysed between copper 
plates there is no back E.M.F. of polarisation, since the 
surface of each plate remains unchanged and the energy 
change at the cathode is compensated by that proceeding at 
the anode. 

The power used in chemical decomposition is obtained in 
watts by multiplying back E.M.F. by current, that is power 
= el watts. 

E e 
. For total I = ~ and power applied = El watts 

where E = applied voltage and I = current in amperes. 
Power spent in heat = PR watts 
Power spent in chemical work = x watts 

= El - PR = El - 


= EI -El +e 

= el watts. 


Overvoltage is another factor which acts against the 
impressed electromotive force during electrolysis. It is 
defined as the excess of reverse potential given by an element 
during its deposition over that given by the pure element in 
the same electrolyte. Overvoltage plays an important part 
in electrolytic reduction and oxidation, and will be referred 
to again. It increases with increase of current density, 
according to Tafel, 1 and probably that is why the Overvoltage 
of hydrogen at smooth platinum is greater than at platinised 
platinum. In the former there is less surface and therefore 
a greater current density. 

Overvoltage varies with different electrodes ; for hydrogen 
the values are 

Platinised platinum .... '005 volt. 
Smooth .... '09 

Nickel -21 

Tin -53 ,, 

Zinc. . .70 

Mercury . . . . . . 78 

Electrical Osmosis is the term applied to the transport of 
the constituents of an electrolyte through a diaphragm when 
one is used to separate anode from cathode. There is a 
general movement towards the cathode of ions, dissolved 
molecules and solvent. 


In a suspension, or colloidal solution, the suspended par- 
ticles or colloids are transported to one of the electrodes. 
This has been utilised in the tanning of skins in which the 
tannin colloid is forced towards the cathode, into the skin 
which is placed in a position between the two electrodes, and 
the rate of tanning is accelerated. 

The electrical dehydration of peat has been accomplished 

by placing the air-dried material between a solid anode and 

a cathode which is perforated ; the water is forced through 

the cathode, which acts as a filter, and a peat can be obtained 

1 Zeitsch. phys. Chem.^ 1905, 50, 641. 


in which the water content has been reduced from 60 per 
cent, to 25 per cent. 

Positive colloids travel to the cathode ; such are : 
Fe(OH) 3 , Zr(OH) 4 , Ti(OH) 4 , Cd(OH) 2 , As(OH) 3 , Cr(OH) 3 , 
Ce(OH) 3) Th(OH) 4 . 

Negative colloids which are driven to the anode are : 
Au, Ag, Ir, Pd, Se, Te, S, Si(OH) 4 , molybdic, tungstic, and 
vanadic acids, sulphides, gums. 

Under a pressure of I volt, colloidal silver travels 3*5 
microns * per second. Gelatin and albumen move less quickly. 

Hydrogen and hydroxyl ions (acids and bases) influence 
the direction in which a colloid travels. Albumen moves to 
the anode in alkaline solution, and to the cathode in acid 
solution, and it remains indifferent to the current in neutral 


It is now a common practice to add colloidal substances 
to the bath for electro-plating, and in refining metals by 
electrolysis, in order to improve the nature of the deposited 
metal. A. G. Betts uses *oi per cent, of gelatin in the fluo- 
silicate method for refining lead, and gelatin is also used in 
the perchlorate lead process. A smoother and more coherent 
cathodic coating is obtained than when no gelatin is used. 2 

It has been suggested that this beneficial action of colloids 
is due to their reducing action since reducing agents have so 
often been applied, e.g. hydroquinone, resorcinol, amino- 
phenol ; but certain oxidising substances act equally well, 
sodium nitrate and chloride, and also certain inorganic salts 
which are neither reducers nor oxidisers. 

E. B. Spear found that many inorganic salts improved the 
deposition of copper, and his theory is, that the colloidal 
particles of copper are dissolved by the addition of a salt 
which increases the solvent power of the electrolyte for the 
metal, and so prevents these particles from depositing and 
keeps the deposit smooth. 3 

1 A micron = 'ooi mm. 2 Met. and Chem. Eng., 1912, 10, 298. 

* Int. Cong. Applied Chem., 1912, Xa, 99. 


According to Mueller, the colloids are protective and 
prevent the colloidal metal from falling out of solution at 
the moment of discharge. 1 


The minimum amount of electricity needed to produce 
one gm. -equivalent of any substance is 96,500 coulombs or 
26'8 ampere-hours. 

In practice, more is needed owing to unavoidable wastage 
and side reactions. 

For example, in the production of chlorine and caustic 
soda it is possible that oxygen may be evolved instead of 
chlorine, owing to secondary reactions. In the manufacture 
of sodium by the Castner process some of the sodium reacts 
with water which is formed during the electrolysis, and 
further, some of the discharged sodium dissolves in the 
fused caustic, travels to the anode by diffusion and is there 
converted to oxide. 

In the manufacture of hydrosulphite of soda, the substance 
itself is unstable, and decomposes to some extent. In this 
case the loss is kept down by increasing the current concen- 
tration, that is the ratio, current / volume of electrolyte. 2 

Evidently it is impossible to obtain a current efficiency of 
loo per cent, in most processes. The efficiency is given 
by the ratio, yield actually obtained / yield calculated from 
Faraday's law. 

In copper refining, current efficiency is very high, about 
95 per cent. The efficiency in alkali-chlorine cells varies 
from 50 to 95 per cent., and for the Castner sodium cell it is 
about 45 per cent. 

The energy efficiency must always be taken into account, 
because it includes not only the current, but voltage also, and 
it is obtained by multiplying the current efficiency by the 
ratio, calculated voltage / actual voltage used. 

With the electrolysis of aqueous sodium chloride the 
current efficiency may be 90 per cent. The voltage may be 

. Elektrochem., 1906, 12, 317. 
3, 2212. 


as low as 23 volts, but sometimes reaches 4 volts; in 

latter case the energy efficiency of the process is ^~ 

= 517 per cent. 


A technical bath can take up to about 6 volts between 
anode and cathode ; higher voltage is liable to produce shunt- 
current losses and excessive heating. 

A number of tanks are generally arranged in series 
because the current derived from the D.C. dynamo may be 
at a P.D. of 150 volts, and this must be distributed over the 
units, so that a pressure of 4 or 5 volts is produced between 
each pair of electrodes. 

It is usual to arrange units in several short series, each 
series being in parallel with the rest (see Copper Refining), 
and a generator of low voltage is used. It is then possible 
to stop a series if any unit goes wrong, without throwing all 
the units out of work, as would be necessary, for instance, if 
a generator of 500 volts were used for 100 cells, all in series. 

The amount of current which a single cell will take 
depends upon 

(1) The working temperature to be maintained. 

(2) Current density to be used. 

(3) Size of the cell. 

If a low temperature is to be maintained, a thorough 
circulation of the electrolyte will be necessary, and the cell 
will be small; if, however, temperature rise is permissible, 
then a large cell, taking more current, may be used. 

The cells are generally rectangular and almost filled with 
anodes and cathodes, which alternate and are placed close 
together to prevent resistance losses. In the parallel system 
the anodes are connected to a common lead, and likewise the 
cathodes. Sometimes a bipolar arrangement of electrodes 
is used, and in this arrangement the two end-electrodes are 
connected to the source of current, while the intermediate 


plates become electrodes by induction, on one side positive, 
and on the other side negative (see Copper Refining). 

Anodes, These t should have a low oxygen and chlorine 
overvoltage, since these gases are so frequently discharged, 
and the material must be resistant to attack by chemical 
action. The substances generally used are platinum, graphite, 
magnetite, lead peroxide,, manganese dioxide and ferrosilicon. 1 

When platinum is used, gauze construction is desirable, to 
diminish cost. Platinum anodes are by no means unattacked 
by chlorine, but when alloyed with iridium a much more resist- 
ant electrode is obtained. 2 Chlorine overvoltage is particularly 
low at graphite, and oxygen overvoltage is much lower at 
nickel and iron than at platinum. 

Iron and nickel are very suitable anode materials when 
oxygen is evolved from an alkaline solution. 

Electrodes of manganese dioxide resist oxygen very well 
in acid solution, as also do Ferchland's lead peroxide anodes ; 3 
carbon electrodes must be prepared with every care if they 
are to prove satisfactory. 4 

Cathodes. These are often of iron or graphite and do not 
have to stand so much corrosive action as anodes. Copper 
gauze is the cathode material in the Hargreaves-Bird process 
for making sodium carbonate. Lead is used for the pro- 
duction of hydrogen and oxygen by the electrolysis of dilute 
sulphuric acid, and mercury forms the cathode when sodium 
is deposited in the electrolysis of brine solutions. 

Generally, when the liquor is alkaline, iron is utilised, and 
when acid is used in the electrolyte, graphite forms a suitable 

Both iron and platinum have low hydrogen overvoltage 
values, hence iron is very useful in alkali-chlorine cells. The 
overvoltage for hydrogen at lead and mercury is very high, 
this is an advantage in the mercury cell for soda, where sodium 
deposition takes place and not hydrogen discharge (q.v.). 

Diaphragms are sometimes essential, in order to separate 

1 Eng. Pat., 9079 (1891). 

2 Zeitsch. Elektrochem., 1902, 8, 149. 

3 Eng. Pat., 24806 (1906). 

4 Met.andChem. Eng., 1913, 11, 242. 


anode and cathode compartments. Asbestos, or some asbestos 
composition, is very widely used for this purpose, and in acid 
liquors, aluminium silicate gives good results. Most clays 
and cements contain too much basic material for them to 
be used in acid liquors. 


Fused salts are good conductors and, by electrolysis, 
yield metal at the cathode and chlorine or oxygen at the 
anode. The chlorides are most suitable for this purpose, 
ZnCl 2 , PbCl 2 , NaCl, CaCl 2 , MgCl 2 , but fused caustic soda or 
potash are used with equally satisfactory results. Probably, 
the ions of the salt or hydroxide are present, but to what 
extent is not known. The conductivity increases with rise 
of temperature, and it is of the same order as for aqueous 
solutions, but higher. 

Faraday's laws of electrolysis hold for molten salts and 
one gram-equivalent of each metal, or radicle, is liberated by 
96,500 coulombs. 1 

The production of " metal fog," during the electrolysis of 
fused salts, was first investigated by Lorenz. 2 If a metal 
be melted under its fused salt, and the temperature raised, 
dark clouds of metal rise which dissolve in the molten salt ; 
on cooling, the clouds or fog settle down and re-enter the 
metal. This sometimes causes a loss of metal during electro- 
lysis if the temperature of the bath is above the melting 
point of the metal which is being deposited. 

The phenomenon is possibly akin to colloidal solution 
since it can be prevented by the addition of certain neutral 
salts to the fused mass, and this addition also has the effect 
of increasing the current efficiency of the process. In the 
case of magnesium production, by electrolysis of fused 
carnallite, it has been shown that what was formerly regarded 
as "metal fog" formation is entirely, or chiefly, due to 

1 Zeitsch. anorg. Chem. y 1900, 23, 255 ; Zeitsch. phys. Chem., 1903, 
42, 621. 

2 Zeitsch. Elektrochem., 1907, 13, 582 ; Zeitsch. phys. Chem., 1911, 
76, 732 ; Trans. Amer. Electrochem.^ 1904, 6, 160. 


the formation of a suboxide 1 of the metal, which must 
be avoided if the maximum current efficiency is to be 

Lorenz has shown that "metal fog" formation is retarded 
by the addition of neutral salts to the fused salt, and in the 
case of lead chloride he made a number of determinations of 
current efficiency showing that the value is improved con- 
siderably by the addition of various neutral salts. 

Ferric chloride exerts a deleterious effect on current 
efficiency 2 in molten electrolytes. 

1 Trans. Amer. Electrochem,^ 1915, 27, 509. 

2 Zeitsch. anorg. Chem., 1903, 36, 36. 



Primary Cells. The first cell, or battery of cells, for 
generating a current of electricity was invented by Volta in 
1800. One form of his battery, the voltaic pile, consisted of 
a series of alternating zinc and copper discs, each pair being 
separated from the next by a piece of moist flannel. In 
another form, the crown of cups, the zinc and copper plates 
took the form of cups, which fitted into each other, and each 
pair was separated from the next by a small vessel of similar 
shape which contained salt water. 

Such a battery provided a large current at low voltage 
or pressure, suitable for electrolysis ; by its means Nicholson 
and Carlisle (1800) succeeded in decomposing water, and, 
within a few years, other workers electrolysed various salt 
solutions and produced various metallic deposits upon the 
negative electrode. Prior to this invention of Volta, the 
frictional electric machine was the only means of producing 
electricity, but such high-voltage electricity was unsuitable 
for electrolysis ; produced at high voltage it could not be 
used for current, but only gave a sudden disruptive discharge. 

Following the discovery of Volta's cell, many primary cells 
were devised, involving the same principle, but designed to 
overcome its chief defects : such were Daniell's Cell, 1836; 
Grove's Cell, 1839; Bunsen Cell, 1843 ; Leclanche Cell, 1868. 

These are termed primary cells because they can be used 
as a primary source of current, but at the present time their 
use is restricted to intermittent work such as telephones, 
electric bells, and the ignition of gaseous mixtures in gas 
engines. A brief description will be given of those in present- 
day use, but it must be remembered that they are not used 

16 . 


for electrolytic work because none of them gives a regular 
and continuous current for very long, and further, the cost 
of working with such t cells, in electrolysis, is excessively high. 
A Board of Trade unit of electrical energy (the kilowatt- 
hour) can be obtained from power stations at one penny or 
less, whereas the cost of generating the same amount of 
energy from a Daniell cell would be about tenpence, and 
from a Leclanche' cell the cost would be about eighteenpence. 
DanieWs Cell. A simple voltaic cell consists of a zinc 
and a copper plate partly immersed in dilute sulphuric acid 
contained in a glass vessel. The difference of potential 
between the two plates is about 1*08 volts, and on joining 
the unimmersed portions of the plates by a copper wire a 
current of electricity is obtained, flowing from copper to 
zinc ; bubbles of hydrogen gas are evolved from the copper 
plate while the zinc plate is rapidly corroded. The energy 
of the cell is supplied by the chemical reaction 

Zn + H 2 S0 4 = ZnS0 4 + H 2 . 

After a short time, the current given by the simple cell 
diminishes, and ultimately, almost ceases ; this is due to the 
bubbles of hydrogen gas collecting on the copper plate, which 
produce considerable resistance, and moreover, give rise to 
a back electromotive force. 

This phenomenon is known as polarisation. In Daniell's 
cell, polarisation is removed by dividing the cell into two 
parts by means of a porous pot and using concentrated copper 
sulphate solution in the outer part of the cell in which the 
copper plate is immersed. The zinc is immersed, as before, 
in sulphuric acid solution contained in the porous pot, and, to 
prevent the zinc from becoming corroded (local action) except 
when current is passing, it is amalgamated by rubbing over 
with mercury. 

The chemical reaction supplying the electrical energy is 
CuSO 4 -f Zn = ZnSO 4 + Cu, and polarisation is avoided by 
depositing copper, instead of hydrogen, on the copper plate. 
The E.M.F. of the cell is 1*07 volts. 

Leclanche Cell. In this cell, a carbon rod forms the 


positive element, packed around with manganese dioxide to 
overcome polarisation. The carbon and manganese dioxide 
are contained in a sealed porous pot, and this pot stands in 
a glass vessel which contains ammonium chloride solution 
in which a zinc rod is immersed. In this cell, any hydrogen 
evolved at the carbon pole is oxidised by the solid man- 
ganese dioxide, and in this way, polarisation is kept down 
sufficiently to render the Leclanche cell valuable for inter- 
mittent work. 1 Its voltage on open circuit is 1*4 to 1*5 volts, 
but it drops rapidly, when the circuit is closed, to ri or 
1*2 volts whilst the normal current of 0*1 to O'2 ampere is 

The various "dry cells" on the market are modified 
forms of the Leclanche", in which pastes are used instead 
of dilute solutions, and the top of each cell is sealed with 
pitch, except for a vent-hole to allow the escape of gas from 
the carbon pole. The porous pot is not needed, and the zinc 
rod is replaced by a cylinder of that metal which generally 
forms the outer case of the cell. 

Lelande Cell. This cell was invented in 1883. There are 
two zinc plates, and between them a copper oxide plate 
which acts as the positive and also as a depolariser; these 
plates are immersed in caustic soda solution and, before use, 
the surface of the copper oxide is reduced to metallic copper. 

There are two or three varieties of the original cell which 
are very efficient. The " Neotherm " cell is of this type, it 
weighs 12 Ib. and will give 150 ampere-hours if discharged 
at the one-ampere rate, the electromotive force is 0*9 to 0*6 
volt when in use. The Edison-Lelande is another much-used 

The energy of all primary cells is supplied at the expense 
of the chemicals which must be replenished; they are, relatively, 
very expensive sources of current, but, prior to the invention 
of the dynamo, they furnished the only source of current 
suitable for electrolysis. 

Fuel Cells. Since only 15 per cent, of the heat energy of 

1 Polarisation in Leclanche cells, Trans. Amer. Electrochem., 1915, 
27, 155. 


carbon is transformed into mechanical energy in the steam 
engine, and not more than 25 per cent, in the gas engine, 
many attempts have been made to construct an electrolytic 
fuel-cell in which the reaction C + O 2 = CO 2 would be so 
utilised that nearly all the heat energy might be transformed 
into electrical energy. Such a cell should give a voltage of 
1*05 volts with oxygen (rc>4 volts with air), and the oxidation 
of i kgm. of carbon would give about 9000 amp.-hours. 

Such a cell must consist of a carbon anode and an oxygen 
cathode separated by a suitable electrolyte ; l if the carbon 
and oxygen are in direct contact, local action will take place 
and no current will be produced. 

The chief difficulties in realising such a cell are : first, 
carbon will not ionise ; second, the common forms of carbon 
are not pure and therefore fouling of the electrolyte will take 
place ; third, there is the difficulty of making an oxygen 

A great deal of ingenuity and thought have been expended 
on this problem, but the solution of it seems, at present, beyond 

Secondary Cells. The lead accumulator is the most im- 
portant representative of this class of cells, in which the 
chemicals used up during discharge can be regenerated by 
passing a reverse current from some other source, usually a 

If two lead plates be used as electrodes in a bath of dilute 
sulphuric acid, the positive lead will become oxidised, on the 
surface, to PbO 2 by the oxygen which is discharged during 
electrolysis. At the negative lead, hydrogen gas is evolved 
and the lead will remain in its original condition. 

After conducting this electrolysis for some time (charging), 
on disconnecting the charging source and joining the two 
plates through a voltmeter, a P.D. will be noticed, of about 2 
volts, flowing from the positive PbO 2 plate to the negative 
lead plate; the hydrogen evolved on the PbO 2 surface, will 
reduce it to spongy lead, while the SO 4 ion, liberated on the 

1 Zeitsch. Elcktrochem. 1894, 1, 122 



FIG. 2. 

other plate, will attack it, forming PbSO 4 . The " forming" of 
the plates is generally conducted after the manner introduced 
by Plante", alternate charging in opposite directions, by which 
means each plate becomes coated with a spongy layer of lead 

which is easily converted to 

_ . k.i 

PbO 2 , or readily attacked by 
the SO 4 ion. 

In an ordinary accumu- 
lator there are many plates, 
say six positives PP, and 
seven negatives NN, arranged 
as indicated, in Fig. 2, to give 
a maximum of surface, and at the same time a minimum of 

During discharge, the anode or positive plate becomes 
superficially converted into PbSO 4 according to the following 

Pb0 2 + H 2 =PbO+H 2 0. (i) 
PbO + H 2 SO 4 = PbS0 4 -f H 2 0. (2) 

The cathode plate of lead becomes also coated with 
sulphate by the following changes 

Pb + O = PbO. (3) 

PbO + H 2 S0 4 = PbSO 4 + H 2 0. (4) 

To avoid the lengthy process of " forming " the plates, 
Faure, in 1888, devised the plan of packing or coating the 
surfaces of the plates with red lead, which on being immersed 
in the acid is changed as follows 

Pb 3 O 4 + 2H 2 SO 4 = PbO 2 + 2PbSO 4 -f 2H 2 O. 

During " charge" the positive, already rich in PbO 2 , has 
its sulphate changed thus 

PbS0 4 + O + H 2 = Pb0 2 + H 2 S0 4 ; 

while at the negative, the peroxide and lead sulphate are 
reduced to lead 

PbS0 4 + H 2 = Pb + H 2 S0 4 . 


The reactions during discharge are those described above, 
(i) to (4), and it is evident that the acid becomes more dilute, 
while during the "charge" reactions, the acid is re-formed 
and therefore becomes more concentrated. When fully 
charged the acid has a density of 1*205 to 1*215, an d when 
discharged the density is between 1*17 and rig. 

It is usual now, to make up the cells with Plante-formed 
positives and pasted negatives, so that heavy discharge may 
take place without risk of disintegrating the positive plates. 

The chemical reactions taking place in the lead accumu- 
lator during " charge " and " discharge " are represented by 
the following equation 


Pb + Pb0 2 + 2H 2 S0 4 ^ 2 PbS0 4 + 2H 2 O. 


The rapid " formation " of accumulator plates is often 
carried out by adding a catalyst to the sulphuric acid. This 
added substance takes many forms, namely, perchlorate, 
chlorate, nitrate, sulphite or acetic acid. 1 If the anion ratio, 
concentration of SO 4 / concentration of, added anion, be not 
too great, the PbSO 4 is formed a very short distance from the 
plate and not actually on the plate, so that the surface-formed 
sulphate is thereby rendered more granular. The same 
principle .has been applied to the electrolytic deposition of 
white lead (q.v.) 

The plates being very close together, it is necessary to 
ensure separation and guard against short circuiting by fixing 
glass, ebonite, or wood separators between them. The nest 
of plates is contained in a vessel of glass, ebonite, or wood 
lined with thin sheet lead. The capacity of an accumulator 
is measured in ampere-hours, and, on an average, a cell gives 
3 to 6 amp. -hours per kgm. of lead (2 to 3 amp.-hours 
per lb.). 

The maximum discharge rate (always marked on the cell 
by the maker) should not be exceeded, and a cell should never 
be further discharged once its voltage has fallen to 1*85 volts. 

Accumulators are generally used as a source of current 

1 Zeitsch. Elektrochem., 1909, 15, 872 ; 1911, 17, 554. 



in electro-chemical laboratories and for experimental work ; 
from 85 to 90 per cent, of the amp.-hours put in may be 
obtained from them on discharge, but usually, not more than 
80 per cent. They constitute a valuable means of storing 
electrical energy which, by the way, is not stored as such, but 
as chemical energy which can be transformed into electrical 
energy as required. They are not used for large-scale electro- 
chemical work on account of their expense and the care needed 
for keeping them efficient. The dynamo is the only current 
generator used for industrial work. 

Thermopiles. These are used to a very limited extent as 
a current source, and depend for their working on the genera- 
tion of a current of electricity 
by the heating of a junction 
of two different metals. 

To produce any apprecia- 
ble current, several junctions 
must be connected in series 
(Fig. 3) and the best results 
are obtained with junctions 
of antimony and bismuth. 
One set of junctions being 
heated and the others kept cool, a thermo-electric current is 
produced and continues as long as the difference in tem- 
perature between the two sets is maintained, and the current 
increases as the difference in temperature is made greater. 
By the use of several hundreds or thousands of junctions, it 
is possible to produce a thermopile, such as that of Gulcher or 
Clamond, capable of furnishing current for electrolysis. This 
is evidently a machine for directly converting heat into 
electrical energy, but since only about I per cent, of the 
energy supplied is converted into electrical energy it is 
decidedly uneconomical. 

A thermopile such as that shown in Fig. 4 has fifty 
elements, costs about 12, and gives 3 amps, at a pressure 
of 3 volts. 

In the Clamond thermopile the elements were iron and a 
zinc-antimony alloy, and with several thousands of such junc- 

FIG. 3. 


tions (6000) heated by coke, an electromotive force of over 
loo volts was obtained.' This thermopile is now obsolete, 
the Gulcher type being the only one employed to any extent. 
The Dynamo. The first magneto-electric machine was 
constructed by Faraday in 1831, and a few years later, in 
1840, Woolrich of Birmingham started the manufacture of 
these machines, but they were not sufficiently developed to 
give large and regular currents until 1867. Since that time 
the dynamo (dynamo-electric machine) has been utilised as 
the chief source of current for industrial electro-chemistry. 

FIG. 4. 

The dynamo used by Elkington at his copper refining works, 
in 1869, attracted considerable attention as a triumph of 
electrical engineering, and during 1870-72, Gramme intro- 
duced important improvements in the construction of dynamo- 
armatures which greatly improved their efficiency. 

The principle of the machine is that of a coil of wire 
rotating in a magnetic field so as to cut the lines of magnetic 
force. If a coil ABCD (Fig. 5) be rotated between the poles 
of a magnet NS, a current is generated in the coil, and if AB 
is rising, the current will follow the direction indicated by the 
arrows. When AB has reached its highest position, that is 
when the plane of the coil is vertical, the rate of change of 


lines of force will be zero and consequently the induced 
E.M.F. will be zero ; then, as AB descends and CD ascends, 
the current induced will flow in an opposite direction, so that 
as the coil rotates a change in the direction of current takes 
place during each revolution. Such a current is alternating, 
and in order to take it from the machine, slip rings are used 
(Fig. 6) which rotate on the same axis as the coil, whilst a 
copper brush, resting on each ring, leads the current to or 
from the external circuit. In order to convert this alter- 
nating current (A.C.) into direct current (D.C.) the split ring 
commutator is used (Fig. 7). 

The two halves of the ring are separated by insulating 
material, but rotate on the same axis as the coil ; the ends 


FIG. 5. 

FIG. 6. 

FIG. 7. 

of the coil terminate on the ring, one end on each half, and 
the brushes are so arranged that the top brush, say, always 
collects current from the rising side of the coil. In this 
way the external circuit is furnished with direct or continuous 

The E.M.F. is proportional to the speed of rotation of 
the coil (armature), to the number of turns in the coil, and 
to the magnitude of the flux in which it turns ; an armature 
therefore consists of many turns, and it is wound on a core 
made up of soft iron stampings in order to provide an easy 
path for the flux. The Gramme ring armature is shown in 
Fig. 8, which depicts the manner in which several coils are 
wound on an iron core, and the manner in which the commu- 
tator xy, consisting of a number of copper bars insulated 
from each other, conveys the current to the brushes. With 
a single coil the current fluctuates, but as the number of coils 


increases the current becomes more steady. Pig. 8 also 
shows the manner in which the field magnet is excited by 
the current from the machine itself (series wound), and if 
the coil rotates in a clockwise direction the current in the 
circuit will take the direction indicated by the arrows. The 
drum armature is the form now generally used, the coils 
being wound on the periphery of a core of soft iron stamp- 
ings, and a study of Fig. 9 will show that this follows, more 
or less, the form of the single coil first described. Each 
coil is wrapped round the drum, and the ends are fixed to 

FIG. 8. 

FIG. 9. 

the commutator. The E.M.F. from a series wound machine 
rises with the current taken from the machine, and as it is 
necessary, for ordinary work, that a dynamo shall give as 
constant an E.M.F. as possible, machines are now either 
shunt wound (Fig. 10) or compound wound (Fig. n). In a 
shunt winding the pressure falls with an increase of load in 
the external circuit, that is with a reduction of resistance, 
because then less current traverses the shunt to produce 
the magnetic field, so that by combining shunt and series 
windings in the right proportion a machine is produced which 
gives a practically constant E.M.F. at all loads. 



The above is a very brief description of dynamo con- 
struction, and for full information a book dealing with the 
subject must be consulted. 

Rotary Converter. This is a dynamo constructed on the 
principle of the single coil arrangement, but, with slip rings 
at one end of the armature to receive alternating current 
(from, say, a municipal supply), and a commutator at the 
other end from which direct current can be collected. The 
machine is driven by A.C., but delivers or generates D.C. 

One of these machines, by the British Thomson-Houston 
Company, is shown in Fig. 12. This 500 K.W. converter 
is used in a large soap and alkali works for electrolytic work. 
The photograph depicts the A.C. and D.C. ends of the 

FIG. 10. 

FIG. ii. 

machine, and the electrolytically operated induction regu- 
lator, seen on the base-plate in the foreground, serves to 
control the voltage. The range of voltage is not great, the 
main purpose of the converter being to convert the A.C. 
supply into D.C. for electrolytic work at a voltage which 
is not capable of any great variation. For example, with 
a 3-phase machine, the ratio of D.C. to A.C. is 

to i, 

that is, the voltage of the D.C. is approximately 17 times 
that of the A.C. 

In the figure, A is a small motor for starting up the 
machine, at B are shown the slip rings at the A.C. end, C 



is the induction regulator by means of which a 3*0 per cent, 
variation in the voltage can be obtained. D is the D.C. end 
of the machine and one of the brush-holders for collecting 
the current from the commutator is just visible. 

Motor Generator. A motor is similar in construction to 
a generator, but receives current which causes it to rotate, 
and it may then be utilised for driving machinery. It may 
be used to drive a dynamo, and such a combination is utilised 
when it is desired to reduce the voltage on supply for electro- 
lytic work. The motor may be a D.C. or A.C. machine 
driven, for example, at 250 volts, and this in turn drives a 
dynamo which gives current at perhaps 5 volts, and when 
motor and dynamo are on the same shaft, the arrangement 
becomes a motor, generator. The generator of Fig. I2a is 
used by a company engaged in the electrolytic refining of 
zinc. The motor A driving the machine, is wound for a 
primary voltage of 11,000 volts and is constructed for 
alternating current. The generator B delivers direct current 
at 220 volts, and the brush and commutator arrangement 
for collecting heavy current is clearly indicated in the 

One more matter should receive notice : When a current, 
developed at a waterfall, is to be transmitted to a station 
several miles distant, the waste of energy converted into 
heat in the conducting wires may be considerable. The 
energy generated at the source will be El where E represents 
voltage and I the current, and the heat produced in the wire 
will be equal to I 2 R. The object is therefore to make I 2 R as 
small as possible. One way is to reduce the resistance of 
the wires (R), but to make conductors of greater diameter 
involves an increase in the cost of metal ; the other method 
is to reduce the current, and since the energy in watts is 
represented by El, it is obvious that if I is reduced, E the 
voltage, must be correspondingly increased. This is not con- 
venient with direct current, but with alternating current it 
is possible to raise and lower the pressure by means of trans- 
formers, based on the principle of Fig. 13. An alternating 
current, produced at the source, is transmitted to the primary 


coil of few turns, and it leaves the secondary coil, of many 
turns, at increased voltage, because 

Pressure in Primary _ Number of turns in Primary 
Pressure in Secondary Number of turns in Secondary. 

The fluctuating current in the primary coil produces fluctu- 
ating flux in the iron core, which in turn produces current in 
the secondary, 

After leaving this step-up transformer, the current is trans- 
mitted to the station at perhaps several thousand volts, and 
before entering the station, where contact with such a current 
would be fatal, it is transformed down to a convenient and 
safe pressure by a step-down transformer with many primary 
turns, and few secondary turns. If required for electrolytic 
work, the A.C. is converted into D.C. by a rotary converter 

FIG. 13. 

or a motor generator. Fig. 13 represents the process of 
transforming down from an A.C. generator A. 

Measurement of the Current. Two kinds of meters are in 
use for the measurement of electrical quantities. The ammeter 
for measuring current strength, and the voltmeter for measur- 
ing the potential difference between two points in a circuit. 
In electro-chemical work it is essential to know, in every 
process, what current is being used, and at what pressure it 
is being delivered. The principles of these two instruments 
are the same. Their construction depends on 

(1) The electromagnetic effects produced on a small piece 
of iron moving in a coil through which the current passes 
(moving iron instruments), or the rotating effect on a coil of 
wire moving in a magnetic field (moving coil instruments). 

(2) The heating effect of the current (hot wire instru- 
ments). The calibration of these instruments will not be 
discussed here as it involves a description of the scientific 


basis on which the various means of measuring electrical 
quantities rests ; such a description is beyond the scope of 
this book, and for full information on the matter a text-book 
on the subject must be consulted. However, the essential 
difference between ammeters'and voltmeters should be under- 
stood. The former are usually low-resistance instruments, 
while the latter (voltmeters) are high-resistance instruments. 

All the current passes through the ammeter, and waste 
of energy will result if the resistance of the instrument is 
appreciable (W = I 2 R), and the current through the circuit 


will be reduced since I = -=^ 


For example, if the current used be 100 amps., and the 
resistance of the ammeter be *i ohm, the power wasted will 
be ioo 2 X'l = 1000 watts, but if the resistance be only *ooi 
ohm the power wasted is only ioo 2 X 'OOi = 10 watts. 

In using a voltmeter to measure the pressure between two 
points of a circuit or between the terminals of an electrolysis 
bath, only a small portion of the current is shunted through 
the instrument, otherwise considerable waste of power will 
take place. 

For example, if a voltmeter be placed across a circuit 
where there is a P.D. of ioo volts, if the resistance of the 
instrument be 200 ohms, the current taken by it (E/R) will 
be '5 amp., and the power used (I 2 R) will be 50 watts. 

If, however, the instrument has a resistance of 10,000 
ohms the current taken (E/R) will be 'Oi amp., and the 
power consumed will be I watt only. 


The energy needed for chemical manufacture by electro- 
lysis is obtained from a dynamo or motor-generator ; it was 
the invention of the dynamo, in 1867, which rendered the 
progress of electro-chemical industry possible. 

To drive the dynamo, a source of power is necessary, and 
cheap power is essential, in many chemical industries, if 
electrolytic processes are to compete successfully with purely 
chemical methods. The power sources available are : water 


power, steam power and gas power, and the cheapest of these 
is water power. Consequently, electro-chemical industries 
flourish in those districts where such power is available. 

Niagara affords the most striking example of successful 
water-power development, the Norwegian waterfalls have 
been utilised to a considerable extent, and on the Continent, 
especially in the neighbourhood of the Alps and the Pyrenees, 
electro-chemical industries avail themselves of cheap water 

There are only two important water powers in the United 
Kingdom : one is at Foyers, in Scotland, where the British 
Aluminium Company utilise the fall from the river Foyers to 
Loch Ness, and obtain a fall of 350 ft. for working the 
turbines ; the other power is at Askeaton, in Ireland, where 
carbide is made. 

Special circumstances may cause the power cost to occupy 
a position of secondary importance. 

For example, in the refining of copper or precious metals, 
the cost of power is relatively a small item, of minor im- 
portance compared with the market value of the products 

Aluminium and sodium can be prepared successfully only 
by electrolytic processes, therefore the cost of power need 
not receive such close attention in these two cases as it must 
in the production of electrolytic chlorine and soda, where 
the industry enters into competition with well-established 
chemical processes. 

The cost of upkeep of a water power installation is 
comparatively small, and the cost is largely dependent upon 
initial capital outlay which, on an average, works out at 10 
per H.P. installed. By allowing 15 per cent, for interest and 
depreciation, and 15^. as working expenses per H.P. year, the 
annual cost of a horse-power is about 45^. The supply at 
Niagara varies from 2 to 4 per H.P. year. 

The cost of steam power in this country is between $ 
and 8 per H.P. year, and when the cost exceeds the latter 
figure it is uneconomic for most electro-chemical industries. 

Gas power -occupies a more favourable position, and with 


producer gas the cost varies from 3 to 3 los. per H.P. year. 
The load-factor of power for electro-chemical work is high 
(90 per cent.), and this implies cheaper power than is usually 
obtained from municipal supplies, where the load factor is 
usually not more than 50 per cent., and sometimes as low as 
20 per cent, when much of the machinery is only in use 
during the early night hours for lighting purposes. 

The cost of electrical energy depends on fuel, operating 
expenses and plant-upkeep, rents, interest, royalties and 
depreciation ; interest and depreciation are generally taken 
together as about 10 per cent, of cost of construction. In a 
paper by E. A. Ashcroft 1 on this subject, the following 
values are given for different power sources per H.P. year 

Water power, 1 los. to 4 (Niagara average, 3 ios. ; 
Sault Ste. Marie, 2 ; Norway, 1) ; gas engine power, 5 $s.; 
oil engine power, $ 8s. ; steam power, 6 gs. 6d. The 
author is a strong advocate of water power, but in the 
discussion which followed it was maintained that steam 
power was to be obtained often at less than 5 per H.P. year, 
that is about \d. per B.O.T. unit. 

In a discussion on power costs in this country, D. B. 
Kershaw 2 quoted the cost of generating electricity by steam 
at -32^. to -64^. per K.W.H. 

The ideal price set down by Professor Donnan was 'id. per 
unit, by means of producer gas plant, with recovery for 
sulphate and tar. This low price of electric power is certainly 
something to be aimed at in this country, but at present '2$d. 
per unit is considered very low. 

The cost in electrolytic work of converting from high 
to low pressure will often have to be taken into account 
when a public supply is used. For example, with a motor 
and generator each having an efficiency of 80 per cent., the 
combined efficiency will be 64 per cent., so that the actual 
cost of the power will be one and a half times as great as the 
supply company's charges. 

In a copper refining or electro-chemical works, running 

1 Trans. Faraday Soc. y 1908,4, 134. 

2 Journ. Soc. Chem. Ind., 1913, 32, 994. 


day and night, the load factor would be high, and cost would 
probably be about '^d. per unit, on an average. 

With coal at 4^. to $s. per ton, with day and night running, 
the cost per unit would probably be in the neighbourhood of 

Although in the United Kingdom there are no important 
water power sources, this must not be regarded as fatal to the 
development of electro-chemical industry. Coal is abundant, 
and by the scientific and economic use of it there seems no 
reason why cheap electric power should not be obtained, 
comparing favourably, as regards price, with water power. 

Cheaper power can certainly be obtained by the erection 
of large central power stations near the coal-fields ; such 
stations would supply the surrounding districts with power as 
is already done, in one or two cases, in the north of England. 

The utilisation of blast furnace gases, for generating electric 
power on a large scale, would certainly provide very cheap 
power. 1 


Several articles on Electro-chemistry and Electro-metallurgy, Electrical 
Review, 1901 to 1902, by J. B. Kershaw. 

A review of Electro-chemical Industry, by J. ^\^?cs\,Journ. Soc. Chem. 
Ind., 1901, 20, 663. 

Niagara as an Electro- chemical Centre, J. W. Richards, Electrochem. 
Ind., 1902, 1, ii,49- 

Technical Electro-chemistry in Russia, Trans. Faraday Soc., 1908, 


Some Electro-chemical Centres, J. N. Pring, 1908. 

Electro-chemical War Supplies, Met. and Chem. Eng., 1916, 14,259. 

Electro-chemical Possibilities of Pacific Coast States, Met. and Chem. 
Eng., 1916, 15, 1 8, 279. 

1 See Coal and its Scientific Use, by Professor Bone in this series of 



THE refining of crude copper by electrolysis was first 
carried out, on a commercial scale, by Elkington at Pembrey 
in South Wales, in 1869. The idea of refining copper and 
other metals by electrolysis originated with Charles Watt of 
Kennington, who, in 1851, was granted a patent (13755) for 
refining various metals and also for producing metals from 
their ores by similar means. 

The method was early recognised as being well suited for 
the production of very pure copper, and the development of 
the industry has been greatly accelerated, since 1890, by the 
increasing demand for high-grade copper for electrical work. 
Though several attempts have been made since 1885 to 
produce the metal from its ores by electrolysis, little success 
has been attained, and crude copper is therefore generally 
obtained by the ordinary metallurgical methods, and the 
product, which varies in purity between 97 and 99 per cent., 
is cast into slabs suitable for the refining tanks. 

The greater part of the World's refined copper is produced 
in America, but a considerable quantity is turned out in the 
United Kingdom and on the Continent. The original refinery 
of Elkington is now worked by Elliot's Metal Company, 
whose output, prior to 1914, was about 7000 tons per annum. 
There are big refineries worked by Messrs. T. Bolton & Sons 
at Froghall, Staffs., and also at Widnes, which produced in 
1914 about 10,000 tons of refined metal per year. 

In principle, the process is the same at all refineries, but it 
differs in certain details. The crude copper is cast into slabs 
(the anodes), about 3 ft. long, 20 in. wide, and f in. to I J in. 

D 33 



thick, which are suspended in a suitable vat where they 
alternate with pure sheet cathodes of electrolytic copper. 
The bath or vat contains, as electrolyte, a solution of copper 
sulphate acidified with sulphuric acid, and the passage of the 
current transfers the copper from anode to cathode, while the 
impurities either dissolve and remain in the electrolyte, or fall 
to the bottom ben'eath the anode where they accumulate, 
forming the anode slimes or sludge. It is essential for suc- 
cessful refining that the crude copper shall be of good quality, 
not less than 97 per cent, copper. On an average, anode 
copper contains 98 per cent, of this metal and the remaining 
2 per cent, consists of arsenic, lead, bismuth, iron, zinc, tin 
and sulphur. The cathode copper, that is, the refined metal 
generally contains 99*93 to 99-98 per cent, pure metal. 

Some analyses of English anode copper are given below 






































I'I 9 




6 1 



Generally, crude copper contains about 30 oz. of silver per 
ton, and -fa oz. of gold. 

An American copper containing more than the usual 
amounts of gold and silver, analysed as follows 

Cu 99-25, Ag -338, Au -ooi, As -033, Sb -054, Pb -009, 
Bi -002, Ni '002, S and Te -008, Oxygen -30. 

This metal contained about 60 oz. of silver per ton and 
\ oz. of gold. The fact that the amounts of arsenic, antimony 
and bismuth are low, is probably due to its production in the 
converter whereby these elements are converted into volatile 
oxides. A few thousandths per cent, of these three impuri- 
ties, it is stated, have a marked bad effect on the conductivity 
of the metal. 

Two systems of refining are in use, known respectively as 



the multiple system and the series system, The difference 
denoted by the names, is shown diagramatically in Figs. 14 
and 15. In the multiple system, which is most widely used, 
anodes and cathodes alternate, the anodes are attached to 
a common lead, that is in parallel, and the cathodes of pure 
copper, about ^V m> thick, are likewise attached to a common 
negative lead (Fig. 14). 

The vats are made of wood and lead- lined. They have 
a capacity of 9 ft. X 3 ft. X 3 ft., and each one takes about 
22 anodes and 23 cathodes, each plate being separated from 
the next by about 2 in. 

The size of the vats must not be greater, on account of 
the difficulty of obtaining efficient circulation of the electro- 
lyte in very large vessels. The voltage applied to each vat is 

FIG. 14. 

about '3 volt with an average current of 4000 amps. Several 
hundreds of these vats (units) are connected in series and 
driven from one generator, the current density (I.D.) is 
generally about 10 amps, per ft 2 , and sometimes is as high 
as 20 amps, per ft 2 ., but if too high, warty masses of copper 
tend to grow across from cathode to anode and produce, a 
short circuit 

The multiple system is very adaptable and is certainly 
preferred for general practice, but the series system is used in 
some large American refineries, i. e, Nichol's Refinery, Brook- 
lyn, and at the Baltimore refinery. In the series (or Hayden 
system) each tank is filled with anode slabs which act as 
bipolar electrodes. On one side pure copper is deposited 
and on the other side crude copper is removed (Fig. 15). 

Each unit carries a smaller current than in the multiple 
system but at a much higher voltage (17 volts). The tanks 
are larger, 1 6 ft. X 5 ft. X 5 ft., and they are made of slate, 


since lead-lined vats would allow leakage of current (at 17 
volts), which would escape via the lining, and current would 
thus tend to pass from end anode to end cathode ; no copper 
would then be deposited on intermediate cathodes. It is 
always found, under ordinary conditions, that more metal 
is deposited on the end cathode than on the intermediate 

The anode copper must be high grade for the series 
system so that the anodes may dissolve regularly, and this is 
further assisted by rolling and hammering. Each unit con- 
tains about 150 bipolar electrodes, placed I to 2 cm. apart, 
each being 6 to 8 mm. thick, and these being thinner than 

FIG. 15. 

multiple anodes, remain in the bath for a shorter time, only 
about 12 days, as compared with 20-24 days. 

The current density used in the series system is high, 
about 2 amps, per dm 2 . (18 amps, per ft 2 .) and hence it is 
necessary to circulate the liquor more rapidly and to regener- 
ate it more frequently than in the parallel system. 

Some advantages of the series system are that for a given 
copper output a smaller plant is needed and less electrolyte, 
and less copper is locked up for a given annual output. 
As the electrodes are closer together, the temperature is 
maintained more efficiently at about 50 C. 

The current efficiency of the multiple system is approxi- 
mately 96 per cent, compared with 90 per cent, for the series 
system, but the voltage drop between each pair of electrodes 
in the former system is nearly twice as great as in the latter, 


25 volt compared with "13 volt; hence the multiple system 

*QO *2 > 

requires ~ x .- - = i'8 times as much energy. One ton of 

refined copper, by the multiple system, requires the expendi- 
ture of 300 K.W.H., and by the series system about 150 

Although the series system shows a great saving in energy 
over the multiple system, and although less metal is " locked 
up " in this process, its successful working is dependent upon 
a regular supply of high-grade anode material of fairly con- 
stant composition. This may not be available, and then the 
cost of preliminary treatment, to prepare anodes of the desired 
quality will probably prove too expensive. The multiple 
system has the great advantage of being more adaptable, that 
is it can deal with low-grade copper and also with copper 
containing extra large quantities of silver or gold. 1 

The American vats are usually larger than those used in 
Europe. They are always supported on insulators of glass or 
porcelain and arranged so that any leakage of liquor can be 
quickly detected. The electrodes are suspended either by 
lugs resting on busbars and the opposite side of the tank, or 
they are suspended by hooks to crossbars. The latter method 
means economy of anode, as the " remainder," which has to be 
re-cast, when the rest of the anode is used up, is reduced from 
about 20 per cent, to 8 per cent. 

Some idea of the scale of working and of the costs involved 
may be obtained from the following statement : A copper 
refining plant using 1000 H.P. per year has a turnover of about 
15,000 tons. The market value of this is approximately 
800,000, and about one-twelfth of this is permanently 
" locked up " (about 67,000 worth), so that the interest charge 
on this will be about 3350, hence the need for driving the 
voltage as high as possible to increase the turnover. 

Whichever system is used, it is necessary to circulate the 

electrolyte and to renovate it at intervals. Circulation is 

necessary to prevent impoverishment of the copper in the 

neighbourhood of the cathode, and renovation prevents too 

1 Eng. and Mining Jour n., 1916, 101, 9. 


great an accumulation of impurities in the electrolyte and 
their consequent deposition on the cathode. 

The upper limit to current density (I.D.) is set by the 
amount of impurities present in the anodes, since the higher 
the I.D. the greater tendency is there for the impurities to be 
carried to the cathode. 

Some form of gravity circulation is generally adopted (see 
Fig. 16), and the rate of flow is governed by the current 
density used. *At Anaconda, where the current density is 
10 amps, per ft 2 ., the rate is three gallons per minute, 
whilst at Great Falls, where a current density of as much 
as 40 amps, per ft 2 , is used, the rate is six gallons per 

The current density is made as high as possible for speed, 

Fro. 1 6. 

but if too high, hydrogen may- be evolved at the cathode, 
spongy copper deposited and even cuprous oxide. Too high 
a current density also tends to dissolve the silver which should 
go into the anode slimes. 

In the Norddeutsche Affinerie, the I.D. is about '4 to '5 
amp. per dm 2 . (3-6 to 4*5 amps, per ft 2 .), and in America, using 
very pure anodes, it goes up to 4 amps, per dm 2 . (36-40 amps. 
per ft 2 .). The average I.D. is 1-2 amps, per dm 2 . (10-20 
amps, per ft 2 .). 

As regards the electrolyte, a high copper content is 
desirable, to keep down the deposition of impurities on the 
cathode and to give a coherent deposit of copper; on the 
other hand, it must not be sufficiently concentrated to permit 
crystallisation of copper sulphate on the anode. Usually, 
about 1 6 per cent, of CuSO4,5H 2 O is present, and 6-10 per 


cent, of sulphuric acid. The acid increases the c6nductivity, 
and prevents the precipitation of cuprous oxide, but if 
present in too great quantity it reduces the solubility of the 
copper sulphate, and causes liberation of hydrogen at the 

Small amounts of sodium chloride, magnesium chloride, or 
hydrochloric acid are sometimes added, to facilitate the pre- 
cipitation of antimony, bismuth, and silver in the sludge. 
Arsenic is so weakly basic that even in an acid solution its 
cathodic deposition is not serious. The addition of a soluble 
chloride causes the removal of antimony and bismuth as 
insoluble oxychlorides. 

The process is continuously controlled by determinations 
of the acidity, copper content, and conductivity. The average 
temperature is 40 ; a higher temperature means increased 
conductivity but tends to produce cuprous oxide. 

The sludge contains silver and gold as well as lead 
sulphate. Most of the antimony, tin, and bismuth go to the 
sludge direct, but to some extent these metals dissolve, form- 
ing unstable sulphates which are hydrolysed, and the basic 
sulphates so formed go to the sludge. 

Arsenic, iron and nickel dissolve in and contaminate the 
electrolyte, whilst cuprous oxide is partly dissolved and has 
the bad effect of neutralising some of the free sulphuric acid ; 
most of it, however, goes into the sludge. 

If the anodes are much below 98 per cent, purity they 
dissolve irregularly and disintegrate, and the electrolyte 
requires more frequent renewal to prevent it becoming foul. 

The chief ingredients of the sludge are generally silver, 
copper, lead and antimony, but other elements are present as 
may be shown by a typical analysis of sludge 

Cu iroi Ag 53-9 Au -29 Pb -91 
Bi -93 Sb 6-25 As 2'i i Se -39 
Te 1-17 SO 4 5*27 H 2 O 2-38 

Analyses showing the difference between refined and anode 
copper are given below l 

1 Zeitsch. Eleklrochem., 1903, 9, 387. 




Cu 99-25 

Pb -009 



Ag '24-'34 

Ni '002 



As -02--03 

Au -ooi 



Sb -007 

Pt -009 



Fe -or 

S,Se,Te -008 



Bi -003 

Oxygen -3 



Oxygen -007 

Analysis of a foul electrolyte (per litre) : Cu 5 1 -80, Fe 1 3-20 
As 14-0, Sb -62, H 2 SO 4 48-0. 

A slime from Anaconda contained the following : Ag 55*15 
Au -2, Cu 13-8, SO 4 10-68, Pb, Bi, Sb, As 8-35. 

These slimes are usually worked up by digesting with hot 
sulphuric acid while air is blown through the liquor to pro- 
mote the solution of copper, arsenic, antimony and bismuth. 
The undissolved portion consists chiefly of lead, gold and 
silver, and is cupelled after the addition of fresh lead. 1 

Selenium and tellurium, if present, are removed by fluxing 
the residue with carbonate and nitrate of soda. The residual 
silver-gold alloy is subsequently cast into anodes for parting 
by electrolysis. 

Foul electrolytes are often regenerated by evaporation and 
removal of the copper sulphate (bluestone) which crystallises 
out. After a second evaporation, the liquor is passed over 
scrap iron to remove any copper remaining in solution. 

At Great Falls refinery, the market for bluestone having 
failed in 1899, the process is now adopted of electrolysing the 
foul liquor with lead anodes, so that most of the copper is 
deposited in a pure state ; 2 or the electrolysis is allowed to 
proceed until not only copper, but arsenic and antimony also, 
are deposited, and the solution is then run into crystallising 
tanks where the sulphates of iron, nickel, bismuth and zinc 
separate. The resulting liquor, which is returned to the cells, 
contains per litre, iioo gms. H 2 SO 4 , As -i, Sb '2, Fe I, Ni 5-3, 
Zn i -5 grms. 

1 Met. andChem. En%., 1911, 9, 417. 

- Met. and Chem. Eng., 1911, 9, 154 ; 1913, 11, 509. 


A scheme has been devised by A. G. Betts, 1 and tested 
experimentally, for fractionally depositing the metals present 
in the slimes resulting from copper and lead refining. 

In 1892 the World's output of electrolytic copper was 
32,000 tons, and in 1902 it amounted to 278,900 tons. In 
the last-named period, the quantity of silver produced was 
27,000,000 oz., together with 346,000 oz. of gold, both of 
these collected from the anode slimes produced during the 
refining of the copper. The cost of refining in 1892 was about 
4 per ton, and at the present time it is about i6s. per ton. 

In 1915 the production of copper, in America alone, was 
about 647,000 tons (Eng. and Mining Journ., 1916, 101, 9). 

The following papers contain interesting particulars of 
copper refining practice 

Copper Refining at Great Falls and Anaconda, Electrochem. 
1903, 1, 416. 

Electrolytic Copper Refining, D. Bancroft, Trans. Amer. Electrochem.) 
1903, 4, 55. 

Electrolytic Refining of Composite Metals (in which a number of 
patents are quoted for separating copper and nickel), T. Ulke, Eng. and 
Mining Jonrn., 1897, 114 ; Trans. Amer. Electrochem., 1902, 1, 95. 

Modern Electrolytic Copper Refining, T. Ulke, Trans. Amer. 
Electrochem., 1903, 3, 119. 


The refining of lead by electrolysis has been carried on 
during the last twenty years, and after varying success this 
method of refining the metal has become firmly established. 
Commercial lead, produced by the process of Parkes or Pat- 
tinson, attains a high degree of purity (99*98 per cent.), so 
that electrolytic refining can only become a commercial 
possibility if a considerable saving of cost is proved. The 
amount of silver present seldom exceeds -01 per cent, (less 
than 4 oz. per ton), and it should be remembered that one of 
the objects of electrolytic refining is the recovery of precious 
metals, together with bismuth, antimony and copper. 

The high chemical equivalent of lead is a point in favour 
of electro-deposition, one faraday of electricity (96,500 

1 Electrochem. Ind., 1905, 3, 141. 


coulombs) deposits 103*5 g ms - of l ea d as compared with 
31*5 gms. of copper. 

Much pioneering work has been carried out in America, 
and considerable progress is due to A. G. Betts, whose process 
is widely used at the present time. 

The methods which have been used, or are now in use, 
are as follows : 

Keith's Process was used in New York for several years. 
The crude lead anodes were suspended in a solution of lead 
sulphate in sodium acetate, and each anode was encased in 
a muslin bag which served to catch the anode slime. The 
anodes had a composition approximately as follows : 

Pb 96-4, Sb i -08, As 1-20, Cu -29, Ag -54, Zn, Fe, etc., -49. 

The process was unable to compete with Parkes's process, 
and its working was discontinued. 

Tommasi Process! The electrolyte in this case is lead 
acetate in aqueous sodium or potassium acetate. 

Two anodes of crude lead are suspended in the bath, and 
between them, a thin disc of copper or aluminium bronze, 
which forms the cathode, is rotated ; the spongy lead which 
collects on the rotating cathode is removed by scrapers as 
it revolves. Some lead peroxide is always formed on the 
anodes which causes a back E.M.F., and hence increases 
the voltage required for continuous working. It is un- 
certain to what extent this process has been used on a large 

Another process proposed depends upon the electrolysis 
of fused lead oxychloride at about 500 C. The oxychloride 
contains a certain proportion of sodium or potassium chloride, 
and the cell devised by Borchers for carrying out the process 
is shown in Fig. 17. The cell, which is of iron, is divided 
into anode and cathode compartments by the insulated parti- 
tion P. The fused anode lead is fed in on the positive side, 
where it runs down to B, and after being dissolved, is trans- 
ferred by the current to the negative side of the cell and 
collects as refined lead at A. 

1 Practical Electro-chemistry, B. Blount, 1906, p. 89. 


The fluosilicate process, invented by A. G. ifetts, 1 for 
refining lead has met with considerable success since it was 
introduced in 1902. 

It is now used in America and in England. At the 
Canadian Smelting Works, Trail, B.C., each tank gives 
750 Ib. of lead per day with 4000 amps, at *5 volt per tank. 

The electrolyte is a solution of lead silicofluoride in 
aqueous silicofluoric acid (H 2 SiF 6 ), and the lead salt is pro- 
duced by the action of aqueous silicofluoric acid upon white 

H 2 SiF 6 + PbC0 3 = PbSiF 6 + H 2 O + CO 2 . 

The 35 per cent, hydrofluoric acid, obtained from fluor 
spar and vitriol, is allowed to trickle over quartz, and the 

FIG. 17. 

solution from this tank, containing H 2 SiF 6 , then passes to 
another tank where white lead is added ; the electrolyte 
contains 70-80 gms. of lead as PbSiF 6 , and 100 gms. of 
H 2 SiF 6 in the litre. 

The anodes of crude lead are 3 ft. X 2 ft. x I in. thick, 
and alternate with pure sheet lead cathodes T V in. thick. 
Each tank holds 22 anodes and 23 cathodes, the tanks are of 
wood, lined with asphalt, twenty-eight in series, 86 in. x 
30 in. x 42 in. deep, and the working temperature is 30- 
35 C. 

The anodes dissolve in about 8 to 10 days, and their 
average composition is Pb 98*0, Ag '62, Sb *6, Cu '24, As *2O, 
Bi TO, Sn, Fe, Au '05. 

1 U.S. Pat. 713277 (1902). Electrochem. Ind., 1903, 1, 407. 


The electrolyte contains 400-500 gms. of gelatine for each 
thousand kgs. of lead deposited. Betts found this addition 
of gelatine greatly improved the deposit and prevented the 
growth of lead crystals from cathode to anode. He mentions 
in his patent other substances than gelatine, namely, resor- 
cinol, hydroquinone, sulphurous acid, and 0r///0-aminophenol, 
and he ascribes their value to their reducing action (see 
p. 10). 

The amount of lead refined per day in one works is 
between sixty and one hundred tons, at an average cost of 
1.6s. per ton, which is practically the same as the cost of 
refining copper. The current density required is -9 to I amp. 
per dm 2 ., and there is a small loss of H 2 SiF 6 , which amounts 
to between three and five pounds per 1,000 Ib. of lead 

The refined metal has an average purity of 99*996 per 

Perchlorate Process^ In this process, worked by Siemens 
and Halske, the electrolyte is an aqueous solution of lead 
perchlorate Pb(ClO 4 ) 2 , formed by the action of aqueous per- 
chloric acid upon white lead ; the current density is 2 to 3 
amps, per dm 2 . Very good deposits are obtained from this 
bath, and it is recommended for electro-plating with lead. 


Formerly the stripping of tin-plate was the only practical 
application of the separation of pure tin from other metals, 
and Goldschmidt's process has been much used for this 

The scrap iron is packed in large iron baskets or cages 
which are made the anodes; each basket holds 10-20 kgs. 
(20-40 Ib.) of scrap, and six of them are suspended in each 
tank, which has a capacity of 3 cubic metres and contains 
10 per cent, caustic soda. The temperature is maintained at 
about 70 C., and the cathode current density is I amp. per 
dm 2 . ; the tin so obtained has a purity of about 98 per cent., 
with 2 per cent, of iron and lead. Probably, the de-tinning 
1 Trans. Amer. Electrochem., 1910, 17, 261 ; D.R.P., 223668 (1910). 


of scrap iron will be accomplished entirely by the chlorination 
process, which seems likely to replace the electrolytic process 
of Goldschmidt in the future. 

The "stripping" of metals by electrolysis is of some 
importance. For example, it is applied to the removal of 
brass from bicycle frames. 1 

Tin is at present refined by electrolysis at Perth Amboy, 
N.J., by the American Smelting and Refining Co. The 
crude metal is obtained from Bolivian ores, and considerable 
progress in refining has taken place recently. 2 


Pure iron is in demand for use in transformer cores and 
for experimental alloy work. 

At Leipzig it is produced by Langbein and Pfanhauser, 
who use the Fischer process. 

The electrolyte is a solution of ferrous chloride and 
calcium chloride ; the working temperature is 90 C. and 
current density 10 amps, per dm 2 . The metal has an average 
purity of 99*95 per cent. ; one of the best samples gave the 
following result on analysis : Fe 99.986, S *oo8, P '007. 

According to the German patent 126839, Merck uses a 
solution of ferrous chloride at 70 C. 


This metal is refined on a small scale by electrolysis, 
using platinum cathodes. Solutions of the fluoride, fluo- 
silicate and fluoborate give good deposits. 3 


Many attempts have been made to produce pure antimony 
from solutions in acid or in alkaline sulphide. The chief 
difficulty is to produce a deposit free from arsenic, and, 
according to Addicks, many methods have been tried at the 
Raritan Copper Works, without success. 4 

1 Electrochem. Ind., 1904, 2, 8. 

2 Met. and Chem. Eng., 1917, 16, 9. 

3 Trans. Amer. Electrochem., 1914, 25, 297, 319. 

4 Trans. Amer. Electrochem., 1915, 28, 325. 



The refining of silver and gold are closely associated, since 
most of the materials brought to the refinery contain both 

Such materials are : crude gold bullion containing about 
30 per cent, gold and 60 per cent, silver ; silver-gold alloy, 
obtained from the slimes of the copper refinery, which con- 
tains about 95 per cent, silver, 3 per cent, gold, together with 
2 per cent, of copper, bismuth, lead, tellurium and platinum ; 
slimes from electrolytic lead refining, or the rich silver-lead 
alloy which results from desilverising lead, and which con- 
tains about 94-98 per cent, silver, with gold *5 per cent., and 
copper, bismuth, lead 1*5 per cent.; scrap jewellery and old 
plate, which may contain as much as 50 per cent, of copper. 

The older method of parting gold and silver by nitric or 
sulphuric acid has now been supersed2d almost completely 
by electrolysis. 

For refining silver, some modification of the original 
method of B. Moebius (1884) is used, in which crude silver 
anodes are suspended in a bath of silver nitrate solution 
containing nitric acid. 1 

The silver dissolved from the anodes is deposited on pure 
cathodes and the copper remains in solution, but must not 
accumulate beyond a 4 per cent, concentration, otherwise it 
will be deposited on the cathode. 

The vats are made of earthenware or of pitch pine lined 
inside with tar, the dimensions being usually about 12 ft. x 2 ft. 
X 2 ft. Cathodes .and anodes alternate, and the anodes are 
sometimes encased in cotton bags which catch the slimes of 
gold and platinum. The deposited silver has often a crys- 
talline structure and tends to grow across to the anode; to 
prevent this, some mechanical device is introduced by which 
the cathode surfaces can be scraped at intervals, and the loose 
silver then drops to the bottom of the tank. 

In a later form of cell, Moebius 2 made use of a cathode 

1 The Mineral Industry ', 1894, 3, 189; 1895, 4, 351; Trans. Amer. 
Electrochem., 1905, 8, 125. 

2 Eng. Pat. 532209 (1895). 


which takes the form of an endless travelling band of silver 
or silver-faced rubber ; from this band the silver is scraped 
by an endless band-conveyor which carries the deposited 
silver upwards and tips it into a receiving box. 

At the Philadelphia Mint, the Moebius process is used 
with vertical electrodes, which give complete satisfaction if a 
certain amount of gelatine be added to the bath. This 
renders the deposited silver quite coherent, so that there is no 
necessity to resort to the horizontal cathode, and mechanical 
scraping is unnecessary, since the silver deposit is firm. 

Gold, platinum and tellurium are unattacked, and fall into 
the slimes as the anode dissolves away ; the slimes also 
contain PbO 2 , together with tin and bismuth as basic salts. 
If the gold content is above 30 per cent, the anodes retain 
their original form after the silver has dissolved and passed 
to the cathode. 

The electrolyte has approximately the following composi- 
tion per litre: silver I gm., copper 40 gms., nitric acid '12 
gm., and to keep the bath efficient a certain proportion of 
the contents must be drawn off and replaced at intervals by 
a fresh solution containing silver nitrate and nitric acid. 

The anodes are about f in. thick, and remain in the bath 
36-48 hours. The amount of silver deposited per K.W.H. is 
about 2-3 kgs. and the current density used is 7*5-20 amps, 
per ft 2 . ('8-2 amps, per dm 2 .). 

The following is a brief account of the procedure adopted 
at the Philadelphia Mint for bullion refining. 1 The anode 
material contains 30 per cent, gold and 60 per cent, silver, 
with 10 per cent, copper, bismuth, lead and zinc. The elec- 
trolyte contains 3-4 per cent, silver nitrate and 1*5 per cent, 
nitric acid, with one part of gelatine per 10,000 parts of 
electrolyte, and I.D. is -75 amps, per dm 2 ., at a pressure of 
i volt. 

The anodes are 7j in. long, 2\ in. wide, f in. thick, the 

earthenware tanks used are 40 in. x 20 in. x 1 1 in., and about 

40 cathodes, consisting of silver strips -016 in. thick, go to 

each tank, eight of which make a series. The deposited 

1 Electrochem. Ind., 1906, 4, 306. 


silver is crystalline but firm and coherent owing to the 
presence of gelatine in the electrolyte. 

The anodes retain a small quantity of silver which is 
extracted with nitric acid, and the gold is then melted down. 

The Moebius process is used at Pinos Altos for parting 
silver-gold bullion from Mexican ores. 1 

Balbach-Thum Process? This process is used at two 
large American refineries, the Raritan Copper Works and the 
Balbach Works, Newark. The electrolyte contains 4 per 
cent, of silver nitrate and 1-2 per cent, of nitric acid, and the 
solution is electrolysed with a I.D. of 5*5 amps, per dm 2 , at 
the anode, and 2-2*5 amps, at the cathode ; the voltage 
required is about 3*5 volts per cell. Each cell consists of a 
shallow dish of porcelain (Fig. 18) lined with carbon plates 

FIG. 18. 

which form the cathode ; the anode is supported over the 
cathode in a frame, so that, as electrolysis proceeds, a deposit 
of silver collects on the floor of the cell, beneath the anode. 

At the Raritan Works the anode slimes from the copper 
refinery are boiled with sulphuric acid, and some nitrate of 
soda added to accelerate the solution of the copper as 
sulphate. When this solution has been filtered, the slimes 
contain 8-18 per cent, of copper and 40-50 per cent, of silver ; 
they are then melted down on the furnace hearth, and air is 
blown over the molten mass till the silver-gold content 
reaches 98 to 99 per cent. Anodes are cast from this product 
and refined in the Balbach-Thum cell. At this works the 
electrodes are slightly inclined to the horizontal, to assist the 
mixing of solutions of high density formed at the surface of 
the anode. Three cells are connected in parallel, each anode 

1 Eng. and Mining Journ., 1891, 51, 556. 

2 Electrochem. Ind., 1908, 6, 277. U.S. Pat, 58035. 



surface is about 3*5 ft 2 ., and the I.D. used is '40 amps, 
per ft 2 . 

The Dietzel Process! In this process, the anode of silver- 
rich alloy is dissolved by nitric acid in the anode compartment 
of the cell, and the silver is removed from the solution by 
passing it through a separate vessel containing copper turn- 
ings ; after this, the liquor, rich in copper, passes into the 
cathode compartment of the cell where the copper is deposited 
on rotating cathode drums. 

FIG. 19. 

By reference to Fig. 19, the course followed by the liquor 
can be traced. The two rotating drums MM, serve as 
cathodes, and the liquid entering the cell at B gives up most 
of its copper before passing through the diaphragm K to the 
anode compartment where the silver anodes CC are mounted 
on glass or porcelain insulators. Silver, copper and base 
metals dissolve, gold and platinum are not attacked. The 
liquor leaves by the pipe D and flows into the vessel E 
which contains scrap copper ; here it deposits its silver and 

1 Zeitsch. Elektrochem., 1899, 5, 81. 


flows over to the pressure vessel F, from which it is pumped 
up to A, whence it flows to the cathode compartment of 
the cell. 


The anode material generally contains about 95 per cent, 
of gold and 5 per cent, of silver, with small amounts of base 
metals. Such, for example, is the composition of the anodes 
left after refining a silver-gold bullion, and if the original 
bullion contain as much as 30 per cent, of gold, the anodes 
remaining when the silver has been removed retain their 
original form and can be immediately used for gold refining. 

The process used is due to Dr. Emil Wohlwill 1 of the 
Norddeutsche Affinerie, Hamburg, and the rights to use the 
process were purchased by the Philadelphia Mint in ipoi. 2 

The electrolyte is 2-10 per cent, hydrochloric acid con- 
taining 2' 5 to 6 per cent, of gold chloride, and the temperature 
used is 60-70 C. Current density varies with the silver 
content and must be lower as the amount of silver increases ; 
in other words, the purer the gold anode the higher the I.D. 
that can be used. The maximum I.D., with anodes contain- 
ing 10 per cent, of silver, is 7-5 amps, per dm 2 ., but it is 
generally less, about 6'5 amps, with a silver content of 5 per 
cent. ; anodes must be periodically scraped to remove the 
silver chloride which coats them. 

The essential part of Wohlwill's patent is the addition of 
hydrochloric acid to the neutral gold chloride, to prevent 
chlorine evolution which always takes place when gold 
anodes are used in neutral gold chloride solution. This 
chlorine evolution is objectionable and it represents waste, 
since it should dissolve gold from the anode when working 

Wohlwill found that hydrochloric acid represses chlorine 
evolution and increases the dissolution of the gold anode. 
His experiments were started in 1874, and ultimately he 
established the following facts 

1 Elecirochem. Ind., 1903, 1, 157; 1904, 2, 221, 261. 

2 U.S. Pat, 625863, 625864. 


(1) He proved the suppression of chlorine by addition of 

hydrochloric acid. 

(2) For a given solution there is a maximum current 

density above which chlorine evolution commences, 
but by augmenting the amount of hydrochloric 
acid or by increasing the temperature this I.D. can 
be exceeded with safety. 

(3) A pure platinum anode is not attacked in gold 

chloride acidified with hydrochloric acid, and chlo- 
rine is evolved, but, if alloyed with gold, the 
platinum dissolves. 

The Wohlwill process was slower at first than the older 
chemical method of solution in aqua regia and subsequent 
precipitation of gold by ferrous chloride, but increasing the 
amount of acid allowed the use of higher current density and 
considerable speeding up was possible. 

The plant at the Philadelphia Mint consists of a 5 H.P. 
dynamo furnishing 100 to 600 amps, at 6 volts. The cells 
are of porcelain 15 in. x n in. x 8 in., and seven such cells 
are run in series. Each cell contains 12 amodes and 13 
cathodes in multiple ; the anodes are 6 in. X 3 in. X \ in. 
thick, and fine gold cathodes, y^j- in. thick, are placed between 
the anodes and \\ in. from them ; the usual temperature is 
5-55 C. In 1903 the amount of gold refined per week was 
5000 oz. with the expenditure of r H.P. 

When the solution becomes sufficiently rich in platinum, 
the gold is precipitated by sulphurous acid, and the platinum 
as (NH 4 ) 2 PtCl 6 by the addition of ammonium chloride. 

The limitations of the process, according to D. K. Tuttle, 
are : If the silver exceeds 5 per cent, it will not fall into the 
slimes as AgCl, but will adhere to the anodes and must be 
removed at intervals, and secondly, if the amount of copper 
be excessive the electrolyte requires very frequent renewal. 

The kind of gold worked at Philadelphia is Hong Kong 
gold ; 975 Au, 20 Ag, -5 Pt, -5 Ir. 

Klondike gold : 776-834 Au, 161-219 Ag. 

Bullion of less than 940 fineness is not worked alone but 


blended with a better grade. The refined gold has a fineness 
of 999'8. 

The Wohlwill process was improved, in 1908, by the 
introduction of alternating current. 1 

An alternating current of rather greater (r.m.s.) 2 value 
is superimposed upon the direct current, and this allows a 
considerable increase in I.D. and quicker deposition. 

With 10 per cent, silver content, the I.D. can be raised to 
1 2-5 amps, per dm 2 , if an A.C. which is ri times as great as 
the D.C. be used, and if the ratio AC/DC be raised to r6, 
anodes with 20 per cent, of silver can be treated with the 
same current density. Much higher I.D. can be used with 
anodes containing a normal amount of silver (5 per cent.). 

The slimes consist almost entirely of silver chloride since 
gold only goes to the slimes, in this process, during the early 
stage of the electrolysis. An A.C. dynamo and a D.C. 
dynamo are used in series, and a good deposit of gold is 
obtained even in the cold, whereas with D.C. only, it is always 
necessary to heat the solution to avoid the formation of a 
dark brown or black gold deposit. 

The Cyanide Process (Siemens and Halske Process). 3 
This process is used for recovering gold from very dilute 
solutions in cyanide (3 to 10 gms. per cubic metre), lead 
cathodes are generally employed. The process is much used 
in South Africa. 


Note on the Electro-metallurgy of Gold, W. H. Walker, Trans. Amer. 
Electrochem, 1903, 4, 47. 

Electrolysis by an Alternating Current, J. W. Richards, Trans. Amer. 
Electrochem, 1902, 1, 221. 

Precious Metal Refining at the Geneva Refinery, Met. and. Chem. 
Eng., 1909, 7, 109 ; 1914, 12, 441. 

The Electrolytic Precipitation of Gold, Silver and Copper from 
Cyanide Solutions, C. H. Clevenger, Met. and Chem. Eng., 1915, 13, 

The Refining of Silver and Gold, Met. and Chem. Eng., 1911, 9, 443. 

1 Zeitsch. Elektrochem., 1910, 16, 25 ; Met. and Chem. Eng. t 1910, 8, 
82 ; Electrochem. and Met. Ind., 1908, 6, 450. D.R.P., 207555. 

8 Root mean square. The A.C. ammeter reading gives the virtual 
value, that is, the maximum value divided by N /2. 

8 Electrochem. Ind., 1903, 1, 484 ; 1906, 4, 297. 



MANY attempts have been made since 1880 to obtain 
certain metals from their ores by electrolysis. 

In the processes adopted, the ore is either dissolved in 
some suitable solvent (generally a salt of the metal to be 
extracted plays an important part) or, on the other hand, the 
ore of the metal is made the anode in a suitable bath. 

In the latter case the chief metal will be transferred from 
anode to cathode in the same way that copper is transferred, 
in the refining process, from crude copper anodes. 

Sometimes a fused salt of the metal is used as electrolyte ; 
this is decomposed and the metal discharged at the cathode. 

Considerable success has been achieved in those cases 
where the metals could formerly be obtained only with 
difficulty by chemical methods, e.g. aluminium, sodium, and 
the alkaline earth metals calcium and magnesium. 

A moderate amount of success has attended the efforts 
made to obtain zinc, lead, nickel and copper by this means. 


This metal has been produced in rapidly increasing 
quantities since 1888 by the electrolysis of alumina in 
fused cryolite. Prior to the establishment of this method it 
was obtained in small quantities by Deville's process (action 
of sodium on aluminium chloride), and sold at about i6s. 
per oz. The new process caused the price to fall, in twenty 
years, from 50^. to about is. 6d. per ib. Pioneering work in 
this field was carried out, simultaneously and independently, 
by C. M. Hall in America, and by P. HeVoult in France, 
between 1886 and 1889. Both workers found that when a 
solution of alumina in fused cryolite is electrolysed, the metal 




is deposited at the cathode and oxygen is liberated at the 

The bath at present used is a 15-20 per cent, solution of 
purified alumina in molten cryolite (Na 3 AlF 6 ) which is con- 
tained in a carbon-lined iron tank, and the electrolysis is 
carried out at a temperature of 900-1000 C. This temper- 
ature is derived from the energy of the current, and the 
carbon lining protects the iron of the tank which forms the 
cathode (Fig. 20). The carbon anodes are clamped together 
in rows and dip well into the molten salt, but, of course, must 
not touch the fused aluminium which collects at the bottom. 

Since the melting point of 
the metal is 665 it remains 
in a fluid condition, and can 
be tapped out at intervals. 
Oxygen which is liberated at 
the anodes combines with the 
carbon, to form carbon mon- 
oxide, so that the anodes 
must be renewed somewhat 
frequently. The voltage re- 
quired is about 5*5 volts and 
each unit carries about 10,000 
amps., with a current density 
at the cathode of 100 amps, per ft 2 . ; at the anode the I.D. 
is often as high as 500 amps, per ft 2 . The diagram (Fig. 20) 
shows a section through one of the cells used by the 
British Aluminium Company at Foyers, N.B., where the 
industry was started in 1898. The alumina is decomposed 
during the process and must be charged into the bath 
continuously, but there is no great loss of cryolite since this 
acts only as a solvent. The original cell of C. M. Hall l 
was a carbon-lined iron box 6 ft. x 3 ft. X 3 ft., into which 
four rows of carbon rods dipped. They were 3 in. in diameter, 
15 in. to 18 in. long, and there were forty or fifty of them in 
each cell. 

FIG. 20. 

1 Zeitsch. Elektrochem., 1903, 9, 347, 360; Electrochem. Ind., 1903, 
1, 158. 


Charcoal is thrown on to the surface of the molten elec- 
trolyte to prevent loss of heat by radiation and to protect the 
anodes from corrosion by air at the surface. 

Since 1888 the metal has been produced in large quantities, 
in Switzerland at Neuhausen, in France at Froges and St. 
Michel, in America at Niagara, as well as in Scotland at 

The pure A1 2 O 3 which is required is prepared from 
Bauxite by the Bayer process (1887), which is conducted by 
dissolving the crude A1 2 O 3 in caustic soda and then allowing 
the solution to stand in contact with fresh hydrated alumina ; 
70 per cent, of the A1 2 O 3 is precipitated in this way and, after 
filtering, it is ignited to convert the hydroxide into oxide. 
The melting point of cryolite is 1000 C, and this is lowered 
to 915 by the addition of 5 per cent, of alumina, but accord- 
ing to F. R. Pyne l the addition of 20 per cent. A1 2 O 3 raises 
the melting point to 1015 C. 

Another important point is, that although solid aluminium 
is less dense than solid cryolite, the values for density are 
reversed when the substances are in a molten state, aluminium 
being denser and therefore sinking to the bottom of a bath 
of fused cryolite. 

The following determinations illustrate this important 
fact 2 





Aluminium .... 
Cryolite .... 
Cryolite saturated with A1 2 O 3 



The anodes must be of good quality and not easily 
fractured, since they have to stand a very high current density 
(500 amps, per ft 2 .). Those used by the British Aluminium 
Company are made from petroleum coke. The coke is taken 
from shale retorts, calcined to remove volatile matter, then 

1 Trans. Amer. Electrochem., 1906, 10, 63. 

2 Zeitsch. Elektrochem., 1894, 1, 367. 


ground up and mixed with binding material, and the moulded 
blocks are then kilned in a producer gas kiln. 1 

For every pound of metal separated there is a loss of one- 
half to three-quarters of a pound of anode carbon by oxida- 
tion, an amount equivalent to that required by the equation 

A1 2 3 + 3C = 2A1 + 3 CO. 

The reason that only alumina is decomposed by the cur- 
rent, and not cryolite, is, that the decomposition voltage of 
alumina is considerably lower than that of cryolite, as is 
shown by the following values 

NaF 47 volts, A1F 3 4-0 volts, A1 2 O 3 2-8 volts. 

Alumina is therefore decomposed preferentially, provided 
it be present in sufficient quantity, otherwise, fluorine will be 
liberated ; too high a current density also leads to the decom- 
position of sodium fluoride. The current efficiency of the 
process is high, 75-80 per cent., and according to some 
experiments by F. Haber 2 in which 5-5 volts were used and 
7520 amps., the energy efficiency worked out at about 
27 per cent. The yield in his experiments per 24 hours was 
43* I kgs., and the current efficiency 71 per cent. ; the energy 
consumption per ton of metal is 23,000 K.W.H., hence one 
H.P. year gives about '28 ton. 

Loss is chiefly due to the formation of " metal fog " caused 
by the process being worked at a temperature so much higher 
than the melting point of the metal (see p. 14). 

The following papers contain accounts of laboratory ex- 
periments on the production of aluminium by electrolysis 

de Kay Thompson, Electrochem. Ind., 1909, 7, 19. 
Neumann and Olsen, Zeitsch. Elektrochem., 1910, 16, 230. 
H. K. Richardson, Trans. Amer. Electrochem.^ 1911, 19, 159. 

An exhaustive research on the melting points and densities 
of various mixtures of cryolite, alumina and calcium fluoride 
is described by P. Pascal, Revue de Metallurgie, 1914, 11, 

1 Zeitsch. Elektrochem., 1902, 8, 26, 607. 

2 Met. and Chem. Eng., 1911, 9, 137. 



Since 1885 several attempts have been made to win 
copper by electrolysis of solutions of the ore, or by making 
coarse metal anodes (Cu 2 S,Fe 2 S 3 ) and suspending these in a 
suitable bath of electrolyte. No method is at present working 
with any great success, but much capital has been expended 
on trial plants, and several patents have been taken out. 
The process is a commercial possibility, and it is probable 
that when the mechanical difficulties have been surmounted, 
the electrolytic winning of this metal will become an estab- 
lished industrial process. 

When the somewhat lengthy and complex Welsh process 
for the production of copper is compared with the direct 
electro-deposition of the metal from its ore, the latter process 
is obviously possessed of an attractive simplicity, and many 
efforts have been made to obtain, in one step, a crude copper 
cathode of 97 or 98 per cent, purity which could then be 

The chief difficulties have always been the breakdown of 
diaphragms and the deposit of sulphur which collects on the 
anodes, thereby increasing resistance to a prohibitive extent. 

In Russia, copper has been produced in certain districts 
for some time, from non-pyritic ores of the carbonate or oxide 
class. At Miedzianka, in Russian Poland, and at Karkara- 
linsk in Siberia, the process of Laszcynski and Stager 1 is 
used, in which the carbonate or oxide ores are extracted with 
5 per cent, sulphuric acid, and this solution electrolysed in a 
vat with lead anodes which are encased in flannel envelopes 
to prevent the oxidation of the iron in the electrolyte. 

The first process to be seriously attempted on a com- 
mercial scale was that of Eugenio Marchese 2 (1885) which 
aimed at the separation of copper from sulphide matte 
anodes in a bath of sulphuric acid containing sulphate of 
copper and ferrous sulphate ; the matte had a composition 
corresponding to that of " coarse metal," namely, Cu 2 S,Fe 2 S 3 . 

1 D.R.P., 144282 (1902). 

3 Zeitsch. Elektrochem., 1894, 1, 50. 


Ferrous sulphate was converted to ferric sulphate, by 
anodic oxidation, and this attacked the anode : Cu 2 S + 
2Fe 2 (SO 4 ) 3 = 2CuSO 4 + 4FeSO 4 -f S, the anode was also 
dissolved simultaneously by anodic oxidation. The mattes 
used varied considerably in composition ; one had the follow- 
ing composition : Cu 17-2, Pb 23*7, Fe 29-2, S 21*01, SO 3 70, 
SiO 2 -87, Ag -06. 

Trials on a laboratory scale, at Genoa, gave metal of 
99'95 P er cent, purity ; further trials were conducted at Stol- 
berg and, on a small scale, metal of 99*92 per cent, purity 
was obtained, but a larger plant, to give 500 kgs. per day, 
failed to come up to expectations. Anode sulphur caused 
the resistance to rise very rapidly above the allowable 
maximum, and the anodes dissolved irregularly and crumbled 
away. A few years later attempts were made by the Mans- 
feld Copper Company l to develop the Marchese process by 
using " white metal " anodes (Cu 2 S) ; the same difficulties 
arose with the anodes but apparently were not insurmount- 
able because the process was reported to be running success- 
fully after a twelve months' trial. The chief object in this 
particular instance was to avoid loss of silver during the 
Bessemerising of the matte, by stopping the " blow " before 
the copper content had risen above 76 per cent., as it is found 
that maximum loss of silver, during the Bessemer treatment, 
occurs when the copper has reached 79-80 per cent. The 
" coarse metal " was therefore " blown " in the converter till 
the copper content was 72-76 per cent, and then cast into 
anodes. The voltage was kept down to one volt per bath, 
even with a fairly thick coating of anode sulphur. 

In the Process of Siemens and Halske, 2 the ore is roasted, 
and then extracted in wood troughs fitted with stirrers, when 
the ferric sulphate, formed during roasting, dissolves the 
cuprous sulphide ; Cu 2 S + 2Fe 2 (SO 4 ) 3 = 2CuSO 4 -f 4FeSO 4 
+ S. The solution is then electrolysed in a two-compart- 
ment cell with carbon anodes. After the copper has-been 

1 M&tallurgie, 1908, 5, 27. 

2 Zeitsch. Elektrochem., 1894, 1, 50; Eng. and Mining Journ., 1892, 
53, 327. 



deposited in the cathode compartment, the liquor is trans- 
ferred to the anode compartment where the ferrous solution 
is oxidised to the ferric state and can then be used to extract 
more matte. The voltage required for each cell is about / volt. 

The Hoepfner Process 1 was introduced, about 1890, with 
the object of extracting unroasted ore with a solution of 
cupric chloride, on the basis of the reaction : Cu 2 S -j- 2CuCl 2 *= 
2Cu 2 Cl 2 -f- S. 

The cuprous chloride formed is kept in solution by means 
of brine or calcium chloride, and this is electrolysed at a 
pressure of '8 volt. After depositing most of the copper 
originally extracted from the ore, the liquor is passed to the 
anode compartment, where it is 
oxidised to the cupric state be- 
fore passing to the extraction 
vat which contains the powdered 
ore. The diaphragms frequently 
ca-used a breakdown, and in a 
modified form of the process, 
which has been used with some 
success on the Continent, the 
diaphragm has been abolished. 
The cell used is shown in section 
(Fig. 21). It contains an inlet 
pipe A, for cuprous chloride 

solution which deposits part of its copper as it passes the 
cathode C, on its way to the carbon anode where the cuprous 
solution becomes converted to cupric chloride, and, since the 
latter solution has a greater density than the former, it falls to 
the bottom of the bath beneath the anode and is syphoned off 
by the pipe B. Diaphragms are therefore avoided by taking 
advantage of the fact that the solution of cupric chloride 
has a greater density than that of the cuprous chloride. 2 

It has been proposed to apply the process of Siemens and 
Halske to the water pumped from mines containing sulphide 

FIG. 21. 


1 Chem. Zeit., 1894, 18, 1906. 
Eng. and Mining Journ., 1892, 53, 471 ; Electrochem. Ind., 1903,!, 


ores. Such liquors contain CuSO 4 , H 2 SO 4 , FeSO 4 , Fe 2 (SO 4 ) 3 , 
and, after depositing their copper, could be subjected to 
anodic oxidation and used for extracting copper ores. 1 


This metal is produced in very large quantity by elec- 
trolysis, especially in America. 

Two methods have been in use for some considerable 
time, namely, the electrolysis of solutions of the chloride or 
sulphate, and the electrolysis of the fused chloride. The 
older distillation process is by no means efficient, there is 
considerable loss of heat and of metal, and for this reason the 
electro-thermal smelting process received considerable atten- 
tion prior to 1914. At that time it was regarded with much 
more favour than the electrolytic processes, but recently, the 
position has undergone reversal, and now, while electrolytic 
processes are turning out thousands of tons of metal per 
week, the electric smelting method has receded into the 

Aqueous solutions of sulphate are generally used, and 
these are obtained by leaching roasted zinc ores with dilute 
sulphuric acid. The conditions necessary for the successful 
electro-deposition of this metal from aqueous solutions have 
long been known, and are as follows 

First, moderately strong solutions must be used (40 to 60 
gms. of zinc per litre), because, in dilute solutions, hydrogen 
is evolved at the cathode and the deposited zinc is spongy. 

Second, free acid must be present ('Oi to 'i N.) and good 
circulation of the electrolyte is necessary. 

Third, low temperature is essential since too high a 
temperature produces a spongy deposit. 

Fourth, metals less electro-positive than zinc, must be 
absent, especially copper and arsenic. The last-named 
element causes spongy deposit, and it has been shown that 
as little as '004 per cent, of arsenic causes spongy deposit 

1 Trans. Amer. Electrochem., 1902, 1, 131. See also Electrolytic 
recovery of Copper from CuSO 4 Leach Liquors, with Carbon Anodes, 
L. Addicks, Met. and Chem. Eng., 1915, 13, 748. 


and evolution of hydrogen in a 10 per cent, zinfc sulphate 

The Anaconda Copper Company 1 turn out about one 
hundred tons of electrolytic zinc per day, the metal being 
deposited from sulphuric acid solution upon aluminium 
cathodes. This process has probably been carried out in 
Germany for many years by Messrs. Siemens and Halske, 
in the production of pure electrolytic zinc having a purity of 
99'98-99'99 per cent. At first, anodes of PbO 2 were used, 
but these were found to disintegrate, owing to the high 
current density used, and they were replaced by anodes of 
manganese peroxide which have proved much more durable. 

The Laszcynski process, 2 used in Russian Poland, is of a 
similar nature. Lead-lined vats are employed, and each vat 
takes 1500 amps, at a pressure of 4 volts, the current density 
being about i amp. per dm 2 . 

Messrs. Brunner Mond and Co., at Winnington, Cheshire, 
have used, for some time, a modification of the Hoepfner 
process originally used at Furfurth, in Germany, in 1895-97. 
In the original process, 3 a poor quality blende was roasted 
with 20 per cent, of salt, at 600 C, and the roasted mass was 
lixiviated with water giving a solution containing 10-11 per 
cent, of zinc together with sulphate and chloride of sodium. 
The sodium chloride was crystallised out by cooling to 
5 C. ; iron, manganese and nickel were precipitated by 
bleaching powder solution, whilst lead, copper and thallium 
were removed by addition of zinc dust. The final solution 
contained about 9 per cent, of zinc and 20 per cent, of 
sodium chloride. Carbon anodes were used and diaphragms 
of muslin. 

The process at Winnington utilises the waste chloride of 
calcium resulting from the ammonia-soda works. Oxide of 
zinc is treated with carbon dioxide in a calcium chloride 
solution and the following exchange is brought about 

ZnO + CaCl 2 + CO 2 = ZnCl 2 + CaCO 8 . 

1 Met. and Chem. Eng., 1916, 14, 264 ; 1917, 16, 9 ; 1916, 14, 30, 120. 

2 Zeitsch. Elekirochem., 1909, 15, 456. 

3 Eng and Mining Journ., 1903, 75, 750. 


This chloride solution is then electrolysed with a current 
density of about I amp. per dm 2 , and the metal is deposited 
on rotating cathode drums which press against one another 
to ensure a firm deposit. One ton of zinc requires about 
3500 K.W.H., so that one H.P. year gives about 175 tons of 

Electrolysis of the Fused Chloride. This process has been 
experimentally tested on the small and large scale. Two 
difficulties have been encountered, namely, the wear and tear 
on the cell at high temperatures by external heating, and the 
greater obstacle of dehydrating zinc chloride. 

Lorenz has investigated the production of pure ZnCl 2 very 
completely. 1 

The conductivity of the fused chloride at different tem- 
peratures has been given as : 400 C. '026 ; 450 C. '057 ; 
500 C. *IO4 ; values which are low for a fused salt. Accord- 
ing to Lorenz, the decomposition voltage, at 5oo-.6oo C., is 
1*49 volts. 

Steinhardt and Vogel dehydrated the chloride by evapora- 
tion in vacua. Much experimental work was carried out, in 
London and at Widnes, in conjunction with the United 
Alkali Company, with a view to perfecting electrolysis in 
externally heated cells, but the method seems to have been 
abandoned in favour of heating internally, 2 and the last traces 
of moisture are generally removed by electrolysis. The cur- 
rent efficiency of the process was found to be 91*5 per cent, 
and the energy efficiency was 33 per cent., at temperatures 
between 450 and 500 C. 

In the process of Swinburne 3 and Ashcroft, worked in 
Norway, the blende is mixed with fused zinc chloride in a 
kind of blast furnace, and chlorine is then driven through the 
molten mass to expel sulphur and convert the blende into 
chloride. The fused chloride is then run off and treated with 
zinc in order to precipitate lead, silver and gold, or the silver 
may be extracted by agitation with molten lead, and the 

1 Zeitsch. anorg. Chem., 1899, 323 ; 1900, 284 ; 1904, 461. 

2 Trans. Faraday Soc. t 1906, 2, 56. 
. 3 Eng. Pat., 10829, io829A, (1897). 


fused salt is then ready for electrolysis. The process was at 
one time worked by the Castner-Kellner Company as a 
means of utilising chlorine, the ZnCl 2 produced being sold as 
such, no attempt being made to obtain metallic zinc from it. 

Grunauer 1 found that "metal fog" formation was dimin- 
ished by the presence of alkali chloride in the electrolyte. At 
600 he obtained the following values for current efficiency 

ZnCl 2 73*9-75*9 

+ KC1 92*1-947 

+ NaCl 83-9-89-9 

+ i-2NaCl 89-6-91-2 

A molten zinc cathode can be used in a fireclay bath 
with graphite anodes and the current used, per unit, is about 
3000 amps, at 4 volts. About 10 per cent, of the energy 
is used up at first in electrolysing out the last traces of water 
contained in the molten chloride. 2 

If alkali chloride be used in the fused mass, the resist- 
ance is lowered and less fuming takes place. The energy 
efficiency is about 36 per cent. 

A process for extracting zinc from blende has been pro- 
posed by S. S. Sadtler, 3 wherein the blende is digested with 
caustic soda and hypochlorite ; the resulting alkaline solution 
of the metal is then electrolysed. 

The refining of zinc by electrolysis is not an economic 
proposition except in the case of zinc which has been used in 
Parkes's process for desilverising lead. The alloy contains, on 
an average, 11-12 per cent, silver, 80 per cent, zinc, 6-7 per 
cent, copper, also small amounts of As, Pb, Sb, and Bi, but 
the electrolytic process has not been able to displace the older 
process of distillation. 


The only process which has been working successfully for 
any length of time is that of Pedro G. Salom, for obtaining 

1 Zeitsch. anorg. Chem., 1901, 27, 177. 

2 Zeitsch. anorg. Chem., 1896, 12, 272 ; Electrochem. Ind., 1905, 
3, 63. 

3 Trans. Amer. Electrochem., 1902,!, 141. 


metallic lead from galena. 1 The process has been worked by 
the Electrical Lead Reduction Company at Niagara for some 
years. Powdered galena is packed on antimonial-lead trays 
(Fig. 22), each tray is insulated from those next it by rubber 
rings, and is filled two-thirds full of dilute sulphuric acid, so 

that the powdered ore is 

\: _. - T^~~ ~j c vered by it ; the bottom 

^^^^Sfe^^.^^^*^/^iA' of each tray is in contact 

with the acid beneath it. 
Generally, forty to fifty 
trays are arranged in series, 
and the top and bottom 
ones connected to the 
source of current, so that 

FlG 22 the intermediate trays act 

as bipolar electrodes which 

are negative on the top, or galena side, and positive on the 
bottom. Electrolysis takes place and the hydrogen liberated 
from each tray reacts with the galena forming hydrogen 
sulphide and metallic lead 

PbS + H 2 = Pb + H 2 S. 

The spongy lead produced is usually roasted to oxide 
(litharge), and its average purity is 97 per cent. About 2*5 
volts are used per cell, and one ampere per pound of metal 

Lead is now being separated in the United States by 
roasting galena with salt, after which, the solution obtained 
by leaching with brine is electrolysed. 2 


The Hoepfner process, similar in principle to the Hoepfner 
process for copper winning, was almost the first process to be 
used. Roasted nickel ore was extracted with calcium chloride 

1 Electrochem. Ind., 1902, 1, 18; Trans. Amer. Electrochem., 1902, 1, 
87; 1903,4, 101. 

2 Met. and Chem. Eng., 1916, 14, 30. 


solution containing cupric chloride, and the chief constituents 
of the ore dissolved according to the following reactions 

Cu 2 S + 2CuCl 2 = 2Cu 2 Cl 2 + S. 

NiS + 2CuCl 2 = Cu 2 Cl 2 + NiCl 2 + S. 

This process did not prove a success and it was replaced 
by that of Savelsburg and Wannschaff, 1 in which the matte 
containing about 65 per cent, of nickel and some iron, but 
almost free from copper, is ground with calcium chloride 
solution and treated with chlorine ; nickel and iron dissolve, 
sulphur is liberated and partly oxidised to sulphuric acid. 
After filtering from Fe 2 O 3 , SiO 2 and CaSO 4 , the solution, 
which contains the chlorides of iron and nickel, is maintained 
at 60-70 C. and air is blown in while freshly powdered ore is 
added, from time to time, to precipitate the iron 

- Ni(OH) 3 + FeCl 2 = NiCl 2 + Fe(OH) 3 . 

After decantation, the liquor is electrolysed, using nickel 
cathodes and graphite anodes, with a current density of 1-1*2 
amps, per dm 2 , and a pressure of 4-4*5 volts. The cathode 
nickel has a purity of 99*9 per cent, nickel and cobalt, and 
contains small quantities of iron, copper and silica. 

The Browne 2 process is used by the Canadian Copper 
Company, Brooklyn. The copper-nickel-iron matte, con- 
taining about equal amounts of nickel and copper, is 
desulphurised and one half of the product is cast into 
anodes whilst the other half is granulated and treated with 
chlorine in the presence of brine ; an electrolyte results, 
containing the chlorides of nickel, copper and iron. The 
composition of the anodes averages 54 per cent, copper, 
43 per cent, nickel, and 3 per cent, iron and sulphur. 
Electrolysis is conducted in cement tanks with thin copper 
cathodes ; most of the copper is deposited, the nickel of the 
anodes dissolves and the ratio of nickel to copper in the 
liquor becomes about 80 : I. The remaining copper is 
thrown out by sodium sulphide, the iron is oxidised and 
removed as Fe(OH) 3 , and after concentration nearly all the 

1 Zeitsch. Elektrochem., 1904, 10, 821. 

2 Zeitsch. Elektrochem., 1903, 9, 392. 



nickel chloride crystallises out in a pure state. This is then 
used to make up a bath, and electrolysed with nickel strip 
cathodes. The cathode metal averages, nickel 99'85, copper 
014, iron "085 per cent. 

Nickel is produced by the Orford Copper Company, New 
Jersey, 1 using anodes of nickel matte in a bath of nickel 

Some of the difficulties encountered in the electrolytic 
winning of nickel are described in Metallurgie, 1904, 1, 77, 
also Electrochem. Ind.> 1903, 1, 208. 


This metal has only been prepared in large quantities, 
at a reasonable price, since the electrolytic process was 
introduced, in 1890, by H. Y. Castner. 

The Castner process depends upon the electrolysis of 
fused caustic soda at a temperature near its melting point 
N which varies with the 

Acg^^^Bl^v^ ' =* quality of the caustic. 

';* hSri nSy^&v^xwj 

rure caustic soda melts 

at 327 C., but the com- 
mercial product may 
have a melting point as 
low as 300 C. 2 The 
melting point of sodium 
chloride is high (800 C), 
unfortunately, a fact 
which renders it difficult 
) to construct a durable 
cell for the electrolysis 
FlG> 23> of molten salt, but one 

or two processes have been patented, and used commercially. 
Castner's process is worked by the Castner Alkali 
Company at Niagara, and by the Castner-Kellner Company 
at Wallsend. Molten caustic is kept at a temperature of 
315-320 C., in an iron pot set in brickwork R (Fig. 23), 

1 Electrochem. and Met. Ind., 1906, 4, 26. 

2 Zeitsch. Elektrochem., 1909, 15, 539. 


the heat being supplied by the ring of burners G. The 
anode F is of nickel, and surrounds the cathode which is an 
iron rod H, fixed by solid caustic K, in the base of the pot. 
A metal gauze cylinder M, between the two electrodes, 
guides the liberated sodium D to a cylindrical iron receiver 
C, beneath the cover N, where it collects in an atmosphere of 
hydrogen. A nickel anode is preferable to one of iron, since 
the former metal is less attacked by any chloride present in 
the molten soda. 

The temperature must be carefully controlled for suc- 
cessful working, even a few degrees rise in temperature 
above 320 leads to a very marked fall in the amount of 
sodium collected. 

Small explosions occur at intervals owing to intermixture 
of hydrogen and oxygen which cannot be completely avoided, 
but as the cells are never very large (18 in. diameter and 2 ft. 
deep at Niagara) such explosions are not dangerous. 

Each unit at Niagara 1 takes about 250 Ib. of molten 
caustic and uses 1200 amps, at 5 volts; the current 
efficiency is about 45 per cent., and current density at the 
cathode is about 2000 amps, per ft 2 . English units are 
generally smaller. 

The reactions which take place during electrolysis are as 
follows 2 

(1) 2NaOH = 2Na* + 2OH'. 

(2) 2Na* + 2H 2 O = 2NaOH + H 2 . 

Primarily, the anhydrous hydroxide is electrolysed accord- 
ing to equation (i) but any water present in the caustic 
will attack the liberated metal according to equation (2). 
Fresh commercial caustic gives a little hydrogen, at first, by 
the electrolysis of water present, but after a short time, 
only sodium is discharged at the cathode. 

At the anode the discharged hydroxyl leads to water 
formation and evolution of oxygen thus 

2OH' = H 2 O + O. 

1 Electrochem. Ind., 1902, 1, 14. 

2 Zeitsch. Elektrochem., 1902, 8, 717. 


If the temperature is too high, sodium diffuses to the 
anode and reacts with the water formed there, so that under 
such conditions both hydrogen and oxygen are liberated at 
the anode, and explosions result. Further losses at the anode 
are due to the following reactions, all oxidation effects due 
to elevation of temperature 

(3) 2Na + 2 = Na 2 2 . 

(4) Na 2 O 2 -f 2Na = 2Na 2 O. 

(5) 2Na 2 O -f 2H 2 O = 4NaOH. 

(6) 2Na + 2 = 2Na*. 

Equations (2) to (6) all represent losses in the Castner 
process, but under good working conditions the only con- 
siderable loss is due to equation (2). 

Current efficiency may be as high as 50 per cent., and 
this could be improved by preventing the access of water 
to the cathode and retarding or stopping the diffusion of 
sodium to the anode. A diaphragm suitable for stopping 
sodium diffusion, made of alumina or sodium aluminate, 1 
has been patented which is unattacked by fused caustic 

Besides the small explosions due to the action of diffused 
sodium upon the water formed at the anode, explosions are 
also due to cathodic hydrogen reacting with the air which 
leaks into the cell from above, and others are produced by 
the electrolytic gases evolved from, the gauze screen sur- 
rounding the cathode, which acts as a bipolar electrode. 
The temperature is kept down as much as possible in order 
to suppress diffusion and to prevent the formation of 
convection currents. 

There is a rapid fall in the current efficiency above 
330 C, not entirely due to the increased solubility of the 
metal in fused soda. Potassium can be obtained in better 
yield by electrolysing molten caustic potash, and one reason 
for the improved yield is that the metal is not so soluble 
in the electrolyte as sodium is in fused sodium hydroxide. 

1 Eng. Pat., 14739 (1902). 



The following measurements of V. Hevesy show the solu- 
bilities in the two cases at different temperatures. 1 


Solubility of Na in 
fused NaOH. 


Solubility of K in 
fused KOH. 


25 per cent. 


7 -8-8 -9 per cent. 


io- 1 













Between 320 and 340 C., a 27 per cent, yield of sodium 
is obtained as compared with a 55 per cent, yield of potas- 
sium, and this loss in the case of sodium is partly due to 
its greater diffusivity. The diffusive power of potassium is 
low and almost constant at 300-550 C., whereas the 
diffusivity of sodium rises at 330 and even more rapidly 
at 340. 

The calculated decomposition voltage of sodium hydr- 
oxide is 2*2 volts, so that, assuming a current efficiency 
of 45 per cent., the energy efficiency will be approximately 

45 X 2-2 


= 22 per cent. 

From this may be calculated the electrical energy used 
in the production of one metric ton of the metal 

4'5 X IPO X 96,500 X IQOO 2 = j ! 700 K W H 
45 x 3600 x 1000 x 23 

Therefore, one H.P. year would give "56 ton. 

The Darling Process? In this process, fused sodium 
nitrate is electrolysed, and nitric acid is a by-product ; the 
method is used at the works of Harrison Bros. & Co., Phila 
delphia, U.S.A., and is regarded very favourably in America, 

A two-compartment cell is used to prevent the liberated 
sodium from reducing the molten nitrate to nitrite. Fused 
nitrate is in the anode compartment, and the cathode space 
is filled with fused caustic soda. The (NO 8 ) ions, discharged 

1 Zeitsch. Elektrochem., 1909, 15, 531. 

' 2 Electrical World, 1902, 39, 136. Eng. Pat., 23755 (1899). 



at the anode, decompose into NO 2 , and oxygen which 
passed into water form nitric acid thus 

HO + 2NO 


HNO 2 + NO 2 = HNO, 
NO + 

HNO 2 , 

= NO 2 . 

The construction of the cell is shown in Fig. 24. The 
inner cathode cell 17 is of perforated sheet iron ; it is placed 

FIG. 24. 

within a perforated iron cell 13, and the space between 
these is filled with Portland cement and magnesia 19, which 
acts as a diaphragm. This central structure is placed within 
an iron vessel 7, which is the anode and contains the fused 
nitrate. The bottom of the anode vessel is covered, to a 
depth of 6 in., with cement, and the innermost cathode vessel 
holds an iron tube cathode 22. Each cell takes 400 amps, 
at 15 volts and a plant of twelve cells decomposes 800 Ib. 
of nitrate per day. Five per cent, of the current is shunted 


through the diaphragm by the switch 31, connected to the 
iron cylinder 17, as this arrangement is found to prolong 
the life of the cell. 

Becker's Process^ In this process a mixture of molten 
caustic soda and sodium carbonate is electrolysed at about 
550C. The cell is similar in design to that of Castner, 
but no wire gauze curtain is used round the cathode, and 
a truncated cone collector is placed over the cathode, and 

FIG. 240. 

connected electrically with it ; this prevents, to some extent, 
the re-solution of the sodium. 

Leblanc and Carrier 2 have investigated the process and 
conclude that it is only a modification of the Castner process 
and not an improvement. 

Contact Electrode Process (Rathenau and Suier.}* The 
process is used by the Griesheim Elektron Co. at Bitterfeld. 
A large iron anode is placed in the middle of a shallow 
bath of caustic soda and is surrounded, at a suitable distance, 

1 Eng. Pat, 11678 (1899). 

2 Zeitsch. Elektrochem.) 1904, 10, 568. 

3 D.R.P., 96672 (1896). 


by a ring of metal cathodes which make contact with the 
electrolyte at the surface only. The current density used 
is about 1000 amps, per dm 2 . 

This process is also intended to be used for the produc- 
tion of the alkaline earth metals, by electrolysis of the fused 

Borchers Cell for Sodium?- This cell is designed for 
electrolysing fused sodium chloride. It consists of a large 
U-tube, the two limbs of which are joined by water-cooled 
connections W (Fig. 240). The smaller limb C is of iron, and 
forms the cathode in which the liberated sodium floats on 
the surface of the fused salt and overflows continuously from 
the outlet pipe. The larger limb is of earthenware and 
contains a carbon anode A at which chlorine is discharged ; 
the gas leaves the anode compartment by the pipe P. Solid 
salt is added at intervals to the chamber which communi- 
cates with the anode compartment and keeps it charged 
with molten salt. 

Ashcroft Process? Invented by E. A. Ashcroft, the 
process is one for obtaining sodium by the electrolysis of 
fused salt, using a molten lead cathode. In one of his 
papers, Ashcroft draws attention to the following facts which 
should render his process of considerable value. The chief 
use of metallic sodium is for the manufacture of peroxide 
and cyanide. During 1906, the United States produced 
1 200 tons of the metal and equal quantities were accounted 
for by the factories of England and Germany. This total 
of 3600 tons was disposed of as follows : for cyanide, 
1500 tons; for peroxide 1500 tons; sold as metal, 500 tons. 
The price of the metal has fallen during recent years from 
2s. 6d. per Ib. to u. 

The patents of H. Y. Castner expired, in England, in 
1905 and apparently there would be less incentive to try 
new processes, but Ashcroft calculates that his process, by 
using fused salt, would almost halve the cost of production 

1 Elektrochem. Zeitsch., 1903, 9, 207. 

2 Trans. Amer. Electrochem., 1906, 9, 123, 362, 355 ; Electrochem. 
hid., 1906, 4, 477 ; Met. and Chem. Eng., 1911 9 253. 



by savings on material and labour. The patent rights of 
the Ashcroft Process are said to have been secured by the 
United Alkali Company, Liverpool. 

The salt is maintained in a fused state at about 780 C. 
in the decomposition vessel A (Fig. 25), which is of iron, 
lined with magnesia bricks. 

The fused lead-sodium alloy forms the cathode, by induc- 
tion, since the anode is inserted in the electrolyte in A. 

At the bottom of the anode cell A, is a layer of molten 
lead which forms the cathode, and the resulting sodium alloy 
is transferred to the decomposing vessel B, through the 
connecting pipe. 

In B, the alloy becomes the anode, in a vessel containing 
fused caustic soda, and through the bottom of which passes 

FIG. 25. 

an insulated nickel cathode. The temperature of the fused 
salt in A is about 770 C., and in B the temperature is 330 C. 

Air is excluded from B, and the sodium which rises to the 
top overflows into T. No hydrogen is evolved at the cathode, 
so that the current efficiency is doubled as compared with 
a process where hydrogen is evolved, and the continuous 
supply of sodium metal without any explosion is facilitated. 

The anode current density is about 2000 amps, per 
ft 2 ., and the lead-sodium alloy is circulated by the action of 
a magnetic coil placed in A. The receiving vessel T is 
terminated inside the cathode cell by an inverted funnel 
which is placed vertically over the nickel cathode so that 
it receives the liberated sodium. 

Another process using fused salt is that of Seward and 
V. Kiigelgen. 1 

1 Eng. Pat., 11175 ( I 9 I )- 



The Processes for manufacturing Metallic Sodium, J. W. Richards, 
Trans. Amer. Electrochem., 1906, 355. 

The Electrolytic Production of Sodium, C. F. Carrier, Junr., Electro- 
. Ind., 1906, 4, 442, 475. 


The metal was first prepared by Bunsen by the elec- 
trolysis of fused chloride. At the present time it is being manu- 
factured by the electrolysis of fused carnallite (KCl,MgCl 2 ) 
probably in a similar manner to calcium, but it is difficult to 
obtain details of the procedure. 

The melting point of magnesium is 633 C. and the 
chloride melts at 708 C. Anhydrous carnallite of course 
melts below this, and the working temperature seems to be 
about 650 C. At this temperature, as Oettel has shown, the 
current efficiency is highest, but he recommends a temperature 
of 700-750 to prevent the metal from solidifying. This 
temperature is also recommended by Borchers. 1 

Since sodium displaces magnesium from its fused salts 
the decomposition voltage of NaCi will be greater than that 
of MgCl 2 . At 700 it is 3'2 volts, so that magnesium chloride 
should be decomposed below 3'2 volts. Borchers found that 
5-8 volts were needed, and Oettel used 4-8 volts, so that on 
an average about 6 volts must be supplied. The current 
efficiency obtained by Oettel was 75 per cent., and the energy 

ij r ^/ 3 *2 

efficiency could therefore be ^ = 40 per cent. 

One kg. of metal requires about 177 K.W.H., and the 
addition of some calcium fluoride to the bath makes the 
magnesium globules coalesce better. 

If the voltage be too high, potassium is discharged at the 
cathode. The Aluminium and Magnesium Gesellschaft, 
Hemelingen, 2 use a bath of carnallite containing enough 
sodium chloride to give an equimolecular mixture of the 
three chlorides. During electrolysis, fresh magnesium chloride 

1 Zeitsch. Elektrochem., 1895, 1 3^1, 394- 

2 Zeitsch. Elektrochem., 1901, 7, 408. 


is added and the bath is kept basic by the addition of alkali ; 
the addition of calcium fluoride is found also to have a 
beneficial effect. 

Hohler states that the optimum temperature is 750- 
800 C., when a current efficiency of 70 per cent, can be 
obtained, working with a cathode I.D. of 27-30 amps, per dm 2 . 

According to Lorenz, 1 the contact electrode method is 
used for producing magnesium. 

Magnesium has been produced in very large quantities 
since 1914, and doubtless some modification of the early 
carnallite process is in use. Of two recent patents, 2 one is 
based on the electrolysis of MgCl 2 with CaCl 2 and CaF 2 , the 
other upon the electrolysis of MgO in molten magnesium 


Electrolysis of fused calcium chloride is the basis of all 
modern electrolytic methods for producing calcium. Pure 
calcium chloride melts at 780 C. and the metal itself at 
800 C. By the addition of impurities the fusion point of the 
chloride may be lowered to 750 C. The finely divided metal 
burns in air at 800 C., and the molten metal shows a strong 
tendency to form " metallic fog," hence the range of tempera- 
ture for working is small ; generally 780-800 C. is adopted. 

Borchers and Stockem 3 obtained small quantities of the 
metal by electrolysis, in 1902. Similar success was attained 
by Ruff and Plato, 4 in the same year, by using a bath made 
up with 100 parts of calcium chloride and 16*5 parts of calcium 
fluoride, melting at 660 C. They worked at 760 C. and the 
I.D. was 3-5 amps, per mm 2 , at the cathode. 

The calculated decomposition voltage for CaCl 2 is 3-24 
volts, but the actual voltage needed is much higher, about 
30 volts. 

At the present time a contact electrode process 5 is in use, 

1 Zcitsch. Elektrochem., 1901, 7, 252. 

2 U.S. Pats., 900961, 800489 (1908). 

3 Zeitsch. Elektrochem., 1902, 8, 757. 

* U.S. Pats., 806006 (1902).' Ber., 1902, 35, 3612. 
5 Zeitsch. Elektrochem., 1904, 10, 508. Eng. Pat, 20655 
U.S. Pats., 864928 (1907); 880760 ([908). 

7 6 


by which commercial calcium is produced. The main parts 
of one cell are shown in Fig. 26 (cell of Seward and 
von Kiigelgen), from which it will be seen that the stick of 
metal attached to the iron tongs can be drawn out of 
the electrolyte gradually, so that it always makes contact 
with the fused salt. Current density used is about 100 amps, 
per cm 2 ., and temperature is kept at 780-800 C. The cell 
A is of cast iron, the cathode of iron B is fixed in the 
bottom, and a carbon lining C concentric with the cathode, 
forms the anode. Insulating material separates the electrode 
from the tank, and a cold water-jacket keeps a protec- 
tive layer of solid electro- 
lyte on the bottom of the 
cell. The water-cooled ring 
b serves to collect the 
globules of metal together 
and to form a cover of solid 
metal, the nucleus of the 
cylinder which is to be 

Apparently, the cell with 
contact cathode proves more 
satisfactory as used at Bit- 
terfeld ; no cooling is then 
needed and the calcium stick 
is formed, without much trouble, after starting it on an iron 
cathode cylinder. 

Small scale investigations which indicated by their 
results the best conditions for an industrial process were 
as follows 

Wohler 1 in 1905 used an externally heated iron vessel 
with an iron rod cathode, the '^electrolyte was 100 parts CaCl 2 
and 17 parts CaF 2 , melting at 660 C. The working temper- 
ature was 665-680 C., and I.D. varied from 50 to 250 amps. 
per dm 2 . It was found advisable to continually raise the 
cathode in order to prevent the formation of "metal fog." 
The voltage used was about 38 volts. 

1 Zeitsch. Elektrochem., 1905, 11, 612. 

FIG. 26. 


Goodwin 1 used a bath of fused CaCl 2 at a 'temperature 
just above 800 C. He also found that regular raising of the 
cathode was necessary. 

Tucker and Whitney found the current efficiency about 
60 per cent. 

Frary and Tronson 2 used anhydrous CaCl 2 , without the 
addition of fluoride, and with a voltage of 18-31 volts obtained 
a current efficiency of 80 per cent. Energy efficiency was 

therefore x 3 -4 _ IO p er cent., and hence i kg. of metal 

would require 

IQOQ X 2 x 96,500 X IPO X 25 = 42 K W H 

40 x 80 x 3600 x 1000 

It has been shown that by employing too low a I.D. or 
too low a temperature spongy metal is produced ; on the 
other hand, too high a temperature brings ?bout re-solution 

of the metal. 


Like sodium, this metal is prepared by electrolysis. It 
has been obtained on a small scale by electrolysing a fused 
mixture of lithium and potassium chlorides. 3 

The metal can also be deposited by using pyridine solu- 
tions of the chloride. 4 

Patten and Mott 5 have patented a process for electrolysing 
lithium chloride dissolved in various organic solvents. 

Ruff and Johannsen 6 recommend fused lithium bromide 
with 10-15 per cent, of chloride, using carbon anodes and 
iron cathodes. From such a mixture they obtained an 80 
per cent, yield, with 10 volts and 100 amps. 


The process of Siemens and Halske depends on the 
electrolysis of alkali thioantimonate solution. A two-corn- 

1 Journ. Amer. Chtm. Soc., 1905, 27, 1403; 1906, 28, 85. 

2 Trans, Amer. Electrochem., 1910,^18, 117 125; Journ. Ind. and 
Eng. Chem., 1910, 2, 466. 

3 Compt. rend., 1893, 117, 732. 

4 Journ. Phys. Chem., 1899, 3, 3602. 

5 Journ. Phys. Chem., 1904, 8, 153. 

6 Zeitsch. Elektrochem., 1906, 12, 186. 


partment cell is used, with an iron plate cathode immersed 
in the thioantimonate solution ; the liquid in the anode 
compartment is sodium chloride. 

In Borcher's process, stibnite is digested with aqueous 
sodium sulphide until the density is 12 Baume, and the 
solution is then electrolysed in iron tanks which form the 
cathodes. Lead anodes are used and a voltage of 2-2-5 volts 
per cell. 

The slimes obtained from lead refining by the fluosilicate 
method, contain 3-4 per cent, of antimony, and this could 
probably be extracted as thioantimonate and then deposited 
by electrolysis. 

A. G. Betts 1 has proposed the use of acid solutions con- 
taining iron salts, for example, SbCl 3 and FeCl 2 , or antimony 
fluoride with ferrous sulphate. During electrolysis the ferrous 
salt is oxidised at the anode to ferric salt, and this can be 
used to extract more crude antimony or antimony ore. 


Crude bismuth contains about 94 per cent, of the metal 
with lead 2*2 and silver 3*1 ; the remainder is composed of 
copper, antimony and gold. It was proposed by Mohn 2 
(1907) to use this crude metal as anode material in a bath of 
bismuth chloride acidified with hydrochloric acid, but the 
process is not satisfactory since it is impossible to prevent 
deposition of the antimony, copper and silver on the cathode 
together with bismuth. 

A method proposed by Foerster and Schwabe 3 gives 
better results, and a good deposit of pure bismuth has been 
obtained. A solution of bismuth fluosilicate is used, and 
from this electrolyte, bismuth, antimony and lead can be 
separated satisfactorily since the potentials needed for their 
deposition are sufficiently far apart. 

1 Trans. Amer. Electrochem., 1905, 8, 187. 

2 Electrochem. Ind. t 1907, 5, 314. 

3 Zeitsch. Elektrochem., 1910, 16, 279. 



THE production of these gases by electrolysis has been 
an industrial operation since about 1895, and there are 
now several different plants on the market for carrying out 
the process. They are found chiefly in accumulator works, 
where the oxy-hydrogen flame is used for lead welding, and 
in works where a high temperature is needed for melting 
refractory metals such as platinum. 

The development of electrolytic hydrogen and oxygen 
plant has therefore been largely dependent upon the ex- 
tended use of the oxy-hydrogen- flame. To produce either 
of these gases, for storing and transport, by electrolysis 
involves competition with well-established and cheap pro- 
cesses such as the liquefaction of air process for oxygen and 
the numerous processes at present available for the cheap 
production of hydrogen gas. 

Cheap power would enable the electrolytic process to 
compete, under favourable conditions, but such conditions 
are not possible at present, in countries where water power 
is not available. 

Before considering the various processes in use, the funda- 
mental data connected with the electrolysis of water will be 

The decomposition voltage may be calculated from the 
heat of formation of water which is 68,400 calories, therefore 
the electrical energy necessary to decompose one gm.-mole- 

cule of water (H 2 O) will be 68)4QQ = 285,714 volt-coulombs 

or joules, since I joule = '239 calorie. 



Since two equivalents of hydrogen will be liberated, which 
require 2 x 96,500 coulombs, the decomposition voltage will 

be _-.._ 8 

2 X 96,500 

In practice, the calculated voltage cannot be realised and 
the pressure required for large-scale work varies about 1*9-4 

Since water itself is almost a non-conductor, it is neces- 
sary to add acid or alkali in order that electrolysis may 
take place ; the solutions used therefore, are dilute sulphuric 
acid 10-20 per cent., or dilute soda or potash 10-25 per 

The experimentally determined minimum voltage neces- 
sary for the decomposition of these aqueous solutions between 
platinum electrodes approximate very closely to 1*67 volts. 

Since one gram of hydrogen is liberated by the passage 
of 96,500 coulombs, therefore, one amp.-hour (3600 coulombs) 
will give '0374 gm. of hydrogen, that is, -0148 cub. ft. at 
N.T.P. 1 The average current used in an industrial unit is 
400 amps., and this will give approximately 400 X "0148 = 
5 '93 cub. ft. of hydrogen per hour, and simultaneously 2^96 
cub. ft. of oxygen. This amount of gas will be given by 
400 x 1*67 = 668 watt-hours or '668 K.W.H., hence i K.W.H. 
will give 8'8 cub. ft. of hydrogen at a voltage of 1*67 volts. 
In practice 4*5 to 8*25 cub. ft. are obtained per K.W.H. 

The following calculation will show the energy consumed 
per year in an ordinary installation for producing about 
15,170 cub. ft. of hydrogen per day 

The annual production, working 24 hours per day and 
300 days per year, will be 15,170 X 300 = 4,551,000 cub. ft. 

Each cell requires per hour 400 amps, at 2 volts, that is *8 
K.W.H., and if there be one hundred cells in the installation, 
the energy used up per year will be 100 X '8 X 24 x 300 = 
576,000 K.W.H. To this must be added about 25 per cent. 

1 N.T.P. stands for normal temperature and pressure, namely, 'the 
temperature of o C., and the pressure of a column of mercury 760 mm. 
high, at latitude 45, and at sea level, the temperature of the mercury 
being o C. 


for loss through the motor generator, making a total of 720,000 
K.W.H. per year. 

If the gases are to be compressed, then 300 Ib. per sq. in. 
for hydrogen will require about 1*6 K.W.H. per hour, that is, 
r6 X 24 x 300 = 11,500 K.W.H. per year. 

The oxygen is generally compressed to 1800 Ib. per 
sq. in., and this will need 4-8 K.W.H. per hour, that is, 
4-8 X 24 X 300 = 34,500 K.W.H. per year. 

Total energy consumption will be 720,000+11,500+ 
34,500 = 766,000 K.W.H. per year, in the production of 
4i5 51,000 cub. ft. of hydrogen and 2,275,000 cub. ft. of 

A brief description of the earlier forms of apparatus will 
now be given, leading up to the modern forms of plant for 
producing the gases by electrolysis. 

In 1885 D'Arsonval, of the Royal College of France, 
invented apparatus for supplying oxygen by electrolysing 
30 per cent, potash solution. A perforated iron cylinder, 
which was enclosed in a woollen bag, served as anode, 
whilst a corresponding iron cylinder served as cathode. The 
hydrogen was allowed to escape, and the apparatus furnished 
100 to 150 litres of oxygen per day. 

In 1888 Latchinoff of Petrograd, devised the first ap- 
paratus for preparing and collecting both hydrogen and 
oxygen. 1 The electrolyte was 10 per cent, caustic soda, and 
iron electrodes were used, while in another form 10-15 per 
cent, sulphuric acid was the electrolyte in which carbon 
cathodes and lead anodes were immersed ; asbestos cloth 
diaphragms were employed. 

On a larger scale he used an iron tank in which were a 
number of sheet-iron bipolar electrodes, separated from each 
other by parchment sheets. Latchinoff was the first to use 
bipolar electrodes for the electrolysis of water, and the first 
to arrange for the compression of the evolved gases. 

In 1890, Colonel Renard of Paris, Commander of a bal- 
loon corps, produced hydrogen in a cylindrical iron vessel 
which acted as cathode, and in which was suspended a 
1 Elektrochem. Zeitsch., 1894, 1, 108. D.R.P. 51998- 



cylindrical iron anode surrounded by an asbestos sack 
diaphragm. 1 

The apparatus gave 250 litres of hydrogen per hour and 
the electrolyte was caustic soda solution. 

In 1893 Bell's apparatus was patented, but it seems not to 
have passed laboratory size. 2 

In the year 1899 the first modern plant was introduced by 
Dr, O. Schmidt, 3 and was built up on the filter-press principle. 
Reference to Fig. 27 will explain the various parts, construc- 
tion, and mode of working of the Schmidt plant, which is 
manufactured by the Machinenfabrik, Oerlikon, Zurich. 

The filter-press is made up of bipolar iron electrodes 

he o 

FIG. 27. 

separated from each other by diaphragms of asbestos bound 
with rubber edges. 

The iron plates e t which act as bipolar electrodes, have 
thick edges or rirrfs, so that when near together there is a 
cavity between two adjacent plates. This cavity contains 
caustic potash or potassium carbonate solution and is divided 
into two equal parts by the diaphragms d the rubber edge of 
which serves to insulate two adjacent plates from each other. 
Two holes in the thick rims of the iron plate at the top h o, 
and at the bottom w w', communicate through the series so 
that there are two channels above serving to convey away the 
hydrogen and oxygen, while the two channels below serve to 
supply the apparatus with electrolyte. For example, w and // 

1 La Lumiere electrique, 39, 39. 

8 D.R.P., 78146. 

8 D.R.P., 111131. Zeitsch. Elektrochem., 1900, 7, 296. 


are connected throughout with the cathode spaces, w' and o 
with the anode spaces. The two channels for supplying 
water are connected with a main pipe W, and, at the other 
end, the two gas channels communicate with the washing 
chambers H and O ; the stopcock a, is for emptying the 
apparatus. A complete Schmidt apparatus is shown in 
Fig. 28. 

The electrolyte is a 10 per cent, solution of potassium 
carbonate. The voltage required is about 2 f 5 volts and the 
energy efficiency is approximately 54 per cent. 

The purity of the oxygen is 97 per cent, and that of the 
hydrogen 99 per cent., a purity which is sufficient for most 
industrial purposes. If the oxygen is required for medical 
use, it is purified by passing 
over platinum at 100 C, after 
which treatment the oxygen 
content is 99^8 to 99^9 per cent, 
with -i to -2 per cent, of CO 2 
and nitrogen. 

The standard types of plant 
on the market are for working 
with 65 or no volts. Each 
K.W.H. decomposes 134 c.c. 
of water, and this loss must be 

continuously made good so that gases may not collect in the 
electrolysis chambers. 

The cost of a Schmidt outfit for generating 33 cubic 
metres of oxygen per 24 hours is about 6000. For imme- 
diate use, without compression, the gases can be made at a 
cost of t>d. or 6d. per cubic metre. 

When sulphuric acid is used (20-30 per cent.) with lead 
electrodes, the conductivity of the electrolyte is higher than 
that of alkali, but the overvdltages produced at the electrodes 
are greater than those produced at iron. Although 10-25 per 
cent, alkali has a lower conductivity than an acid solution, 
it is possible to use iron electrodes in alkaline solutions, a 
convenience generally utilised in the construction of modern 
electrolytic plant. The greatest difficulty met with in 

FIG. 28. 

8 4 


designing plant for the electrolytic production of hydrogen 
and oxygen, is that of providing a diaphragm which will 
effectively prevent the mixing of the two gases to form an 
explosive mixture. 

Schools Plant? The anodes and cathodes are of lead, 
and each electrode is encased in a glass or earthenware tube 
which is perforated around its lower portion and sealed at 
the top with insulating material. Each electrode is thus 
completely separated from the rest, and mixture of the gases 
is rendered impossible. 

Dilute sulphuric acid (density 1*235) is the electrolyte 


FIG. 29. 

used, and the voltage for each unit is about 3*9 volts. The 
attached diagram will explain the construction of the cell 
(Fig. 29). 

Each unit consists of a cylindrical lead-lined vat which 
contains two cylindrical lead anodes and two corresponding 
cathodes. Each lead electrode contains a bundle of lead 
wires which give a large surface, and the lower "part of the 
electrode is perforated to give free access to the current and 
the electrolyte. The surrounding glass or earthenware tubes 
are also perforated, round the lower portion, for the same 

Iron electrodes can be used in Schoop's apparatus, with 

1 Journ. Soc. Chem. Ind., 1901, 20, 258. D.R.P., 141049. Electro- 
Chem. Ind., 1902, 1, 297. 


alkaline electrolyte, and then the working voftage is about 
2*25 volts, as compared with 3 '8 volts for sulphuric acid 
(density 1*235) and lead electrodes. 

The following costs are quoted by the makers, for plant 
with acid electrolyte 

One H.P. hour gives 97*5 litres of hydrogen together 
with half this amount of oxygen, or stated in another form, 
one cubic metre of the mixed gases requires 6*2 to 6'8 H.P. 
hours at a cost of $d. to ^\d. With an alkaline electrolyte 
and iron electrodes the cost is considerably less. The oxygen 
has a purity of 99 per cent, and hydrogen 97*5 to 98 per cent. 

Process of Garuti. Introduced in 1893, tne plant of 
Garuti and Pompili has had considerable success. The 
electrolyte is alkaline, and iron electrodes are used. 

Anodes and cathodes are connected in parallel, and the 
diaphragms separating each electrode from its neighbours 
are iron sheets perforated near the lower edge. The arrange- 
ment depends, for successful working, on the fact, first ascer- 
tained by Del Proposto, 1 that if the voltage between the 
electrodes is not above 3 volts, the iron diaphragm between 
them does not become bipolar, and this principle is applied 
in the design of the commercial cell. In the early years 
(prior to 1899) Garuti used lead electrodes and sulphuric 
acid, but this was ultimately abandoned in favour of iron and 
alkali. 2 

The outer box and electrode system are of iron (see 
Fig. 30). The electrodes c are 12 mm. apart, and their lower 
edges are 12 cms. from the bottom of the tank. Each dia- 
phragm partition has a zone of perforations, 4 cms. wide > 
running parallel with the lower edge and about 7*5 cms. above 
it. The anode spaces a open at the top, on one side, into a 
main outlet for oxygen. In a similar manner the cathode 
spaces b open, on the other side, to a common hydrogen 

The hydrogen has a purity of 98*9 per cent, and oxygen 

1 Bull de PAssoc. des Ingen. Electr., 1900, 11, 305. 

2 D Industrie ttectrochimique, 1899, 11, 1 13, Eng. Pats., 23663 (1896) ; 
12950(1900); 2820(1902); 27249(1903), 



97 per cent. The average consumption of energy is 4*17 
K.W.H. per cubic metre of mixed gases, values which corre- 



FIG. 3oa. 

spond with a current output of 96 per cent., and an energy 
output of 57 per cent. 

FIG. 30. 

Fig. 300 is a side elevation of a Garuti cell and Fig. 30 is 
a cross section through A A. 

The cost of a 100 H.P. Garuti plant, comprising 50 cells 
each using about 400 amps., and two gasometers for 
collecting the gases, is about 3000. This is increased to 


^4000 if compression plant is required for compressing the 

It should be mentioned, that on an average, the electrical 
energy required for obtaining 2 cub. metres of hydrogen 
and i cub. metre of oxygen, is 13-5 K.W.H. 

Schuckerfs Process?- The process was introduced in 1896. 
Iron tanks are used to contain the electrolyte which is 15 per 
cent, caustic soda, and the working temperature is 70 C. 
Sheet- iron bells are used to isolate the electrodes and collect 
the gas evolved. 

Each tank takes about 600 amps., and has the dimen- 
sions 26 in. x 18 in. X 14 in., and holds about 50 litres. Each 
pair of unlike iron electrodes is separated by strips of good 
insulating material extending from the top, down about three- 
quarters of the total depth. Between these separating plates, 
and enclosing the electrodes, are the iron bells which collect 
the evolved gas and lead it away. 

The plant is manufactured by the Elektrizitats A. G. vorm. 
Schuckert & Co., Nurnberg, and standard types are supplied 
to take from 100 to 1000 amps. 

The following prices are quoted for a plant giving 10 cub. 
metres of hydrogen per hour 

Electrolyser , .'-.". . 470 

Soda 80 

Insulating materials . . 20 

Scrubbers, dryers, etc. . 50 
Two gas-purifying stoves 

and packing . . . 150 



Two gas-holders . . . 400 
Wooden stages for cells 40 
Compressors . . . . 57 

Water still 40 


The International Oxygen Co., 2 New York. The cell con- 
sists of an iron tank, acting as cathode, and from the cover is 
suspended a perforated inner tank which acts as anode, and 

1 D.R.P., 80504. Electrochem. Ind., 1903, 1, 579 ; Elektrochem. Zeitsch., 
1908, 230, 248. 

* Met. andChem. Eng., 1911, 9, 471 ; 1916, 14, 108. 


which is made of low carbon steel to prevent the formation 
of spongy rust. The anode and cathode are separated by 
an. asbestos sack suspended from the cover. The average 
current, per cell, is 393 amps, at 2*6 1 volts, and the working 
temperature is 30 C. The purity of the oxygen is stated to 
be 98*3 per cent, and each cell gives over 3 cub. ft. per 

A cell somewhat similar to this in design is the Halter 
cell. 1 An iron tank forms the cathode and in this an inverted 
box or funnel-shaped iron anode is suspended, and from the 
edge of this an asbestos sack hangs to prevent mixture of the 

Modern Filter-Press Cells. The filter-press form of cell 
has been developed considerably. The National Oxy-hydric 
Co., Chicago, 2 manufacture a cell of this pattern, in which the 
electrodes are corrugated to increase the surface; they are 
made of a special alloy and heavily nickel-plated. The 
electrolyte is 21 per cent, caustic potash, and asbestos dia- 
phragms are used. The purity of the oxygen is 99-5 per 
cent, and 4 cub. ft. per K.W.H. are given, with twice that 
amount of hydrogen. 

Other plants of this type are : the Siegfried Barth, 
Dusseldorf, Oxy-hydrogen generator ; the L'Oxhydrique 
Franchise generator; 3 the generator of Eycken, Leroy and 
Moritz. 4 

The following cells of various patterns have been patented, 
but most of them have not been utilised industrially to any 
great extent 

The Burdett System, U.S. Pat., 1086804 (1914). 
The Tommasini System, U.S. Pat, 1035060 (1912). 
Fischer, Leuning & Collins, U.S. Pat., 1004249 (1911). 
Siemens & Halske System, La Machine, Vol. V, pp. 7, 33. 
Siemens Bros. & Obach, Eng. Pat, 11973 (1893). 
See also Fr. Pats., 355652 (1905); 198626 (1906). 

1 U.S. Pats., 1172885, 1172887 (1916). 

2 Met. andChem. Eng., 1916, 14, 288. > 

3 Fr. Pat., 459967(1912). 

4 Fr. Pat., 397319 (1908). 



By employing a sufficiently high current density at the 
anode during the electrolysis of dilute sulphuric acid, a 
considerable amount of ozone is mixed with the evolved 

To preserve the anode and increase the yield of ozone 
it is desirable that a water-cooled anode should be employed, 
and as low a temperature as possible, below o C. With 
a I.D. of 80 amps, per cm 2 , and a voltage of 7*5 volts, 
using 15 per cent, sulphuric acid, the evolved oxygen contains 
28 gms. of ozone per cub. metre, that is, a yield of 7*1 gms. 
per K.W.H. 1 

With an alkaline solution, a much smaller quantity of 
ozone is produced. 2 

A method for large-scale production of ozonised oxygen 
has been devised by Archibald and Wartenburg, 3 in which 
alternating current is superimposed upon the D.C. used for 
the electrolysis, and it is found that the amount of ozone is 
two or three hundred times as great as that obtained with 
D.C. only. The increased yield is due to the depolarising 
effect of the A.C. on the anode, and the principle is similar to 
that used in Wohlwill's improved process for gold refining. 4 
In a trial run, a maximum yield was obtained with an A.C. 
of 6 amps, and a D.C. of '25 to I amp. 

1 Zeitsch. anorg. Cnem., 1907, 52, 202. 

2 Zeitsch. anorg. Chem., 1903, 36 403. 

3 Zeitsch. Elektrochem., 1911, 17 812, 

4 Page 52. 



THE electrolytic manufacture of caustic soda, chlorine, 
bleaching liquor, hypochlorites, chlorates and perchlorates, 
has developed from the electrolysis of aqueous solutions of 
sodium or potassium chloride, by modification of the condi- 
tions under which electrolysis takes place. 

If a 10 percent, solution of common salt be electrolysed 
between platinum or carbon electrodes, chlorine gas is evolved 
from the anode and caustic soda is formed in the neighbour- 
hood of the cathode. Since the sodium which is liberated at 
the cathode is attacked by water immediately on its discharge 
and converted into caustic soda, the only element liberated 
at the negative electrode is hydrogen. These changes are 
represented by the following equations 

NaCl = Na* + CK ; Na* + H 2 O - NaOH + H\ 

If anode and cathode be not separated by a diaphragm, 
chlorine will diffuse from the anode and react with the caustic 
soda to form hypochlorite, 

2NaOH + C1 2 = NaCl + NaCIO + H 2 O. 

By electrolysing a cold dilute solution of sodium chloride 
without a diaphragm, formation of hypochlorite results, 
whilst, if the solution be more concentrated (25 per cent.) 
and hot (50 C.), chlorate formation is favoured, especially if 
the solution be slightly acid, 

2HC1O -f- NaCIO = NaClO 3 + 2HC1. 

The chlorate will be converted into perchlorate if the 
current density be high and the temperature low, 

NaC10 3 + 2OH' + 20 = NaCIO, + H 2 O. 


The patent of Charles Watt, 13755 (1851), cohered the 
preparation of chlorine, soda, hypochlorite and chlorate by 
electrolysis of alkali chloride solutions. 

The production of chlorine and caustic soda by electro- 
lysis is a rapidly growing industry and is well established in 
most countries ; its development will be discussed first, and 
then the production of the other valuable, but less widely 
produced, substances will be considered. 


Hydrogen and sodium liberation are both possible at the 
cathode. The hydrogen ions are more easily discharged, 
since a cathode potential of only "4 volt is required, as 
compared with a discharge potential of 271 volts for 
sodium in a solution which is normal with respect to sodium 
ions, therefore, under ordinary conditions hydrogen is liberated 
and caustic soda is formed. 

The electrolytic solution pressure of sodium can be re- 
duced by employing a mercury or lead cathode, when an 
alloy with the liberated sodium will be formed. The discharge 
of the sodium is likewise facilitated by using a cathode of 
metal which has a high hydrogen overvoltage, so that the 
liberation of sodium at the cathode in aqueous solution 
becomes possible. 

Even by raising the concentration of salt it is possible to 
obtain sodium discharge at ordinary temperatures if the ionic 
ratio Na*/H* is thereby made sufficiently high. 

The mercury cell is the only technical unit in which, at 
ordinary temperature, sodium is liberated ; in all others hydro- 
gen is liberated and a cathode metal with low hydrogen 
overvoltage is used. Iron is very suitable, its overvoltage 
value for current densities between I and 10 amps, per 
dm 2 , being -3 to '55 volt. 

The anodes which are used must withstand the action of 
chlorine and are therefore made of platinum, graphite or 

9 2 



Griesheim Elektron Cell? This cell has been much used 
on the Continent. It was one of the earliest industrial 
chlorine-soda cells, and though it is not so efficient as 


(TB c 


FIG. 31. 

some others, it is simple in construction and cheap, gives 
chlorine of good quality and has very little diaphragm 

Each unit consists of a rectangular iron box (Figs. 31-33) 



FIG. 32. 

which is steam-jacketed, and covered with material which 
conducts heat badly. It is mounted on insulating blocks and 
contains six rectangular boxes made of cement, about I cm. 
in thickness. These cement boxes act as diaphragms and 
contain the anodes A. The outer iron box forms the cathode, 
and cathode plates are also provided in the form of iron sheets 
1 Ber., 1909, 42, 2892 ; Chem. Z/., 1909, 33, 299. 


C placed between each anode compartment, and* reaching 
almost to the bottom of the cell (Fig. 31). The cell is shown 
in plan in Fig. 32. 

A cross section of the cell (Fig. 33) shows the arrangement 
of pipes for steam S, salt solution inlet B, outlet for chlorine 
D, and outlet for caustic soda liquor E. The liquor is gener- 
ally run through at such a rate that about one-third of the 
salt used is converted into caustic. Saturated salt solution 
is used and the working temperature is 80-90 C. If the 
caustic is allowed to concentrate beyond the above strength, 
oxygen is given off at the anode by the electrolysis of the 
water, and part of the current is wasted. 

The current density used 
is 100-200 amps, per square 
metre (10-20 amps, per ft 2 .) 
and the pressure for each 
unit is about 4 volts. 

The anodes are of mag- 
netite (Fe 3 O 4 ) made from 
decopperised burnt pyrites 

(Fe 2 O 3 ) which is fused in the 
electric furnace, and a little 
fresh ferric oxide (Fe 2 O 3 ) 
added to obtain a homo- 
geneous mass of Fe 3 O 4 . 

The cylindrical magnetite anodes are said to be much 
cheaper than graphite (about one-fifth the price) and they 
last much longer, moreover, they give purer chlorine. Best 
quality carbon anodes give 5-8 per cent, of CO 2 , and oxygen 
often amounts to 6-8 per cent. 

The cement casing, of which the anode cells are made 
(D.R.P. 30222) is obtained by mixing cement with salt 
solution containing hydrochloric acid, and after setting has 
taken place, the boxes are soaked in water which washes 
out the more soluble constituents including the salt. In 
this way a very porous diaphragm results, which offers 
small resistance to the current and which has proved very 



\ A 


^ C 







i >. 

FIG. 33. 



The best working conditions for this cell may be 
summarised as follows 

(1) Anodes, preferably of magnetite, to obtain pure 

chlorine and complete absence of carbon dioxide 
in the anode gas. 

(2) High concentration of brine. 

(3) Temperature should be high in order to reduce the 

voltage required (80-90 C.). 

Hargreaves-Bird Celll A company was formed in 1900 
(the Electrolytic Alkali Co.) with works at Middlewich, 

Cheshire, for the production of car- 
bonate of soda, using this cell. The 
French rights were purchased by 
the St. Gobain Co., and the process 
is used, in America, by the West 
Virginia Pulp Co. 

Each cell consists of a rect- 
angular iron box, lined with cement, 
or, in place of iron, sandstone blocks 
clamped together form the contain- 
ing vessel. JThis box is about loft, 
long, 4 ft. to 5 ft. deep and 2 ft. 
wide, and is divided into three parts, 
longitudinally, by two asbestos 
sheet diaphragms D. Six carbon 
anodes are placed in the central or anode division, and 
the cathodes C consist of two sheets of copper gauze the 
same size as the diaphragms and attached to them on the 
outside (Fig. 34). The copper matting or gauze, which forms 
the cathodes, is sufficiently strong to form a support to the 
diaphragms and it is therefore more correct to state that the 
diaphragms are attached to the cathodes. The two cathode 
spaces between the copper cathodes and the sides of the cell 
are empty, except for steam and CO 2 which pass into the top 
of the cell, when it is working, by the pipes SS. 

1 Eng. Pats., 18039, 18871 (1892); 5197, 18173 (1893)- Electrical 
World, 1905, 46, 101 ; Journ. Soc. Chem. Ind. (1895), 14, ion ; 
Electrochem. Review, 1900, 20. 

FIG. 34. 


The anode compartment is filled with saturated" brine, and 
during electrolysis this brine percolates through the diaphragm. 
When it reaches the copper cathode sheet, caustic soda is 
formed and this is swept to the bottom of the cell by con- 
densed steam and CO 2 ; it therefore leaves the cell as a 
solution of sodium carbonate or bicarbonate, mixed with 
sodium chloride, by the pipes OO. 

Twelve cells generally run in series, taking 2000 amps., 
that is, a current density of about 20 amps, per ft 2 ., and using 
a pressure of 4-4*5 volts. About 66 per cent, of the salt is 
converted into Na 2 CO 3 , and approximately 240 Ib. of salt are 
converted into sodium carbonate per 24 hours, giving about 
580 Ib. of soda crystals, corresponding to 220 Ib. of calcined 
soda ash. 

The anodes first used were carbon blocks threaded on a 
rod of lead-copper alloy, and where the alloy was left exposed, 
it was packed round with cement. Acheson graphite anodes 
are now used. 

The chlorine produced per 24 hours gives about 360 Ib. 
of bleaching powder, containing 37 per cent, of available 
chlorine. Reference to the diagram (Fig. 34) shows that the 
diaphragms must be about 10 ft. long by 4 to 5 ft. deep, and 
they are about \ in. thick. 

According to Taussig, the Electrolytic Alkali Co. now 
make bicarbonate of soda by this process. 1 

Kellner has devised a cell of the filter-press type for 
making alkali carbonate and chlorine. 2 

Outhenin- Chalandre Cell? This cell, which is used in 
France and Italy, consists of an iron box which is divided 
into three sections by vertical partitions. The outer divisions 
contain the cathode liquor, and the inner one is the anode 
compartment containing graphite anodes immersed in strong 

The diaphragm is made up of a number of cylindrical 
unglazed porcelain tubes which are cemented into the dividing 

1 Trans. Faraday Soc.^ 1910, 5, 258. 

2 Journ. Soc. Chem. Ind., 1892, 11, 523. 

3 Brochet, La Soude Electrolytiqiie, p. 103. 


walls of the cell in a slanting-position to that. They 
traverse the anode chamber and are open at both ends, thus 
connecting the two cathode compartments. Each tube con- 
tains an iron cathode and the cathodes are all connected to 
the negative pole of the current source ; their sloping position 
facilitates the escape of hydrogen. 

A 1400 amps, unit contains 108 cathodes which are 
arranged in six rows, and there are 19 anodes. The voltage 
required is about 4 volts per cell and a K.W. day is reported 
to produce about 67 kgs. of caustic soda. 

The cell is undoubtedly more complicated than the 
Griesheim cell and requires more attention. 

Townsend Cell. 1 This cell is used in America, and 
appears to have worked very successfully. It commenced 
working in 1905 at Niagara with a plant taking 1000 H.P. 
The first cells were designed to take 2000 amps., but the 
later types are capable of using 5000 to 6000 H.P. 

The construction and working of the cell can be best 
understood by reference to Figs. 35 and 36, from which it 
will be seen that a diaphragm is used, and that the structure 
somewhat resembles that of the Hargreaves-Bird unit. How- 
ever, the special point of the Townsend cell is the employment 
of kerosene oil in the cathode compartment to facilitate the 
rapid removal of caustic soda, as formed, from the cathode, 
so that diffused chlorine from the anode has little chance of 
converting it into hypochlorite and chlorate. 

The foundation of the cell is a somewhat massive cement 
foot G (Fig. 35) which is shown in section in Fig. 36, and 
which takes the form of a wide U. The outer containing- 
walls of the cell are made up of two strong iron plates CC 
which are clamped firmly to the cement foot. The bulge on 
these plates forms the cathode chambers and they carry 
kerosene inlets DD, and caustic soda outlets EF. The 
asbestos diaphragm and the cathode of iron gauze are kept 
in position by the same clamps which hold the outer walls to 
the central foot. 

1 Electrochem. and Metall. 2nd., 1907, 5, 209; 1909, 7, 313; Int. 
Cong. App. Chem., 1909, Sect. X. 36. 



The arrangement of diaphragm and cathode is shown in 
the figure, where it will be seen that the 'diaphragm A forms 
the wall of the anode chamber, and the cathode B is close 
against it. B is also in contact with the outer iron wall 
which is connected to the negative pole of the current 
source. The anode compartment contains a hollow graphite 
anode which almost fills the anode space, and through it 
runs a pipe, by which saturated brine is pumped into the 

FIG. 35. 

When the cell is working, the liquor between anode and 
cathode is partly converted into caustic soda which streams 
through the diaphragm to the outer compartment. This 
exterior chamber being filled with kerosene, the heavier caustic 
liquor falls to the bottom, where it collects, and ultimately 
flows out through the outlet pipes EF. 

Each cell is about 8 ft. long, 3 ft. high, and 12 in. wide, 
and is easily taken to pieces for cleaning or repairs. The 
current density used at first was about 100 amps, per ft 2 ., 
but more recently it has been forced up to 150 amps., with 


satisfactory results. The voltage required is about 4 volts per 
unit ; and about 1 5 to 20 litres of brine pass through each cell 
per hour. The resulting caustic soda liquor contains about 
1 50 gms. NaOH and 200 gms. NaCl per litre, but it is possible 
to run the brine through at a rate of 24 litres per hour, and, 
by increasing the current density, to obtain a caustic liquor 
holding 200 gms. of NaOH in the litre. The caustic is free 
from hypochlorite and chlorate, and this is a sign of high 
efficiency because a fall in efficiency is always attended by 
formation of NaCIO and NaClO 3 , and the presence of these 
substances is objectionable on account of their corrosive 

action on the anodes, 
and on the vacuum pans 
during the subsequent 
concentration of the 

Usually, the dia- 
phragm is cleaned every 
30 days, and composite 
graphite anodes are used 
so that only the corroded 
portions need to be re- 
placed. Aluminium con- 
ductors have also been 
as they are more resistant 


d M 

FIG. 36. 

used instead of copper ones 
to the action of chlorine. 

The diaphragm used in this cell is a patent of H. L. 
Baekeland, and consists of asbestos, the pores of which have 
been filled with ferric oxide and colloidal Fe(OH) 3 . 

Current efficiency is 96-97 per cent., and the usual voltage 
is 3*4 to 3*6 volts. The process after a three years' run was 
pronounced to be quite satisfactory. 

By reducing the rate of percolation of the liquor through 
the diaphragm, it is possible to obtain as much as 250 gms. 
of NaOH per litre. The level of the anode liquor controls the 
rate of percolation. 

In Fig. 36 is seen the adjustable glass pipe N, which slides 
up and down in an outer tube L ; this glass pipe communi- 


cates with the anode liquor, and by raising it, the level of 
the liquor is raised, whilst if it be lowered the excess 
of anode liquor runs away through it. In the same figure 
MM are flushing channels for washing out the bottom of 
the cell ; and K is an outlet for chlorine gas from the anode 

The cathode liquor (about 14 per cent. NaOH) runs out 
in two continuous streams from the side tubes EF. The 
loss of oil by evaporation, etc., is small, about eight to ten 
shillings' worth per day in a large plant. 

About 5 tons of caustic were produced daily at Niagara, 
and 1 1 tons of bleaching powder ; the finished caustic 
contains about 2 per cent, of Na 2 CO 3 and a little NaCl. 

The life of the vacuum pans and finishing kettles is said 
to be very long on account of almost entire absence of 
chlorate and hypochlorite. The "bottoms" which are left 
in the finishing kettles are also said to be small, amounting 
to only i ton per 700 tons of finished caustic. 

The energy efficiency of this cell is reported to be 45 per 
cent., a little less than that of the Griesheim cell. 

Finlay Cell. 1 This cell is the subject of Eng. Pat. 1716 
(1906) by Messrs. Archibald and Finlay of Belfast. Accord- 
ing to Professor Donnan, it has given good results and 
produces a caustic soda liquor containing 12 per cent, of 
NaOH. On the other hand, it has been stated that the 
strength of caustic soda is only about 8 per cent., and this 
would detract from the value of the high energy efficiency 
of the cell which is stated to be about 75 per cent. 

The industrial unit is constructed on the filter-press 
principle, and some idea of its construction can be gained 
by a study of Figs. 37 and 38. Fig. 37 gives a section 
through one type of single experimental cell from which 
it will be seen that a sheet-iron cathode is separated from 
the graphite anode by two asbestos diaphragms DD ; 
between these is an intermediate space into which the brine 
solution flows from the cistern B. Outlets are provided in 

1 Trans. Faraday Soc., 1909, 5, 49 ; Allmand, Applied Electro- 
Chemistry, p. 380. 



anode and cathode compartments for the gas and liquor 
produced in each, EA and HC. 

The mixing of anode and cathode liquor is entirely 


E | 

jo OH 

A - 

- A 












i i 






FIG. 37. 

avoided by the motion of the brine from the centre feed 
chamber. This prevents diffusion of chlorine and formation 

C D 

FIG. 38. 

of oxy-halogen compounds, and it also prevents the passage 
of OH' to the anode from the cathode. In the technical 
unit the separate cells are fixed end to end until a unit 
of sufficient size has been built up'; Fig. 38 shows this 


arrangement. Cathode and anode spaces are formed by 
means of distance frames in which the electrodes are sup- 
ported. The iron cathode C, for example, is followed by 
a cathode diaphragm of asbestos D, against this is a waxed 
cardboard separator T \j- in. thick, and this forms the brine 
space enclosed on the other side by the anode diaphragm ; 
then follows the anode. 

In this way a series of anode and cathode chambers are 
formed, with communicating pipes leading to the main pipes 
for outlet of liquor and gas. A 2000 amps, unit only occupies 
5ft. x 2'5ft. x 4ft high, and very pure chlorine is obtained. 

Macdonald Celll This cell is in use at the New York 
& Pennsylvania Company's Paper Mills, Johnsonburg, Pa., 
and the plant there gives 16 tons of bleaching powder per 
day together with 6| tons of caustic soda. 

The cell is also used by the Standard & Colorado City 
Works for chlorinating gold. It is simple in construction 
and gives satisfactory and economical results. Each cell is 
made up of an iron tank which serves as cathode, 5ft. long, 
2 ft. high and 2 ft. wide ; the tank is divided longitudinally 
into three sections by two perforated iron plates. The 
middle portion forms the anode compartment and contains 
ten graphite anodes ; it is separated from the outer cathode 
compartments by asbestos diaphragm sheets which are fixed 
on the inside of the perforated iron plates which divide off 
the anode compartment. 

The main object of the installation at Johnsonburg was 
to provide chlorine for bleaching purposes ; the caustic soda 
was for some time run to waste, but subsequently arrange- 
ments were made for collecting and evaporating it down. 
It contains 3-4 per cent, of salt and 16-18 per cent. NaOH. 

Le Seur Cell? 1 This cell also, has been installed in several 
American paper mills. It is made of iron and is divided 
into two compartments by a diaphragm of asbestos E (Fig. 
39), which is fixed to the iron gauze cathode C. The anode 

1 Electrochem. Ind., 1903, 1, 387 ; 1907, 5, 43 ; Eng. and Mining 
Journ., 1903, 75, 857. 

* U.S. Pat, 723398 (1903)- 



compartment A is filled with brine, and contains a graphite 
electrode; this compartment is sealed by a brine seal HH 
which also serves as the entrance for fresh brine flowing 
from the pipe B, and chlorine escapes through the pipe D. 
The liquid level in the anode compartment is slightly higher 
than in the cathode part to ensure a steady percolation of 
liquid in the right direction, anode to cathode, and the 
caustic which is formed flows out through the outlet K. 
The block G fixes the cathode at its lower end. 

Billiter- Siemens Cell?- 
This is a diaphragm cell 
with a bell anode chamber, 
but the mouth of the bell 
is closed by a diaphragm 
of asbestos which rests on 
the negative electrode of 
nickel netting. Inside the 
bell, the anode of carbon is 
suspended, and the whole is 
immersed in an iron vessel 
which forms the cathode 
chamber. The diaphragm 
is horizontal and it is com- 
posed of asbestos cloth on 
which is laid a powdered 
mixture of barium sulphate, 
alumina and asbestos wool 
which is made into a coherent mass by the addition of salt. 
It is stated that the patent rights have been purchased by 
Siemens & Halske, and the cell is used in several works, 
one of which is the Niagara Alkali Co. 

According to J. B. Kershaw, the cell is one of the most 
efficient of the well known alkali-chlorine cells ; this state- 
ment is substantiated by a study of the efficiencies tabulated 
below. 2 

1 Eng. Pat., 7757 (1907)- Journ. Soc. Chem. Ind., 1913, 32, 993 ; 
Trans. Faraday Soc., 1913, 9, 3 ; Zeitsch. angew. Chem., 1910,28, 1072, 


2 Journ. Soc. Chem. Ind., 1913, 32, 993. 




Current Efficiency. 

Energy Efficiency. 


91 per cent. 

52-3 per cent. 























In these cells a mercury cathode is used which forms 
a liquid amalgam with the discharged sodium ; this amalgam 
is transferred to a separate compartment where it is decom- 
posed by water and the sodium converted into sodium 

Discharge of hydrogen from neutral solution requires a 
cathode potential of "4 volt, and, with the current density 
and temperature usually employed, the hydrogen overvoltage 
at mercury is 1*25 volts which corresponds to a necessary 
voltage of 1-65 volts for hydrogen discharge. 

By using a dilute amalgam the sodium discharge potential 
is lowered from about 2*7 volts to approximately 1*5 
volts, so that, under these circumstances, sodium will be 
discharged in preference to hydrogen. The discharge poten- 
tial for sodium, at the surface of an amalgam saturated with 
that metal, is 1*8 volts. 

The anodes are of platinum or carbon, the working 
temperature is about 50 C., and the amount of sodium in 
the amalgam must not exceed "02 per cent. 

Particles of carbon falling on the mercury surface facili- 
tate evolution of hydrogen, and the chlorine gas therefore 
frequently contains 2-3 per cent, of hydrogen. If platinum 
anodes are used, the amount of hydrogen is usually less 
than -5 per cent. 

The electrolyte used is generally 30 per cent, salt solu- 
tion from which sulphate, lime and iron have been removed. 
Cathode I.D. varies about 5-25 amps, per dm 2 ., and at the 


anodes it is greater ; a high I.D. at the anode favours pure 
chlorine because electrolysis then takes place at the surface 
of the carbon where chlorine concentration is constant. A 
low current density means evolution of oxygen from the 
interior of the anode, and this is accompanied by a certain 
amount of CO 2 . 

Elevation of temperature leads to hydrolysis of the 
chlorine and production of carbon dioxide 

C1 2 + H 2 O = HC1O -f- H* + Cl'. 
The advantages of the mercury cell are 

(1) Concentrated soda liquor can be obtained (24 per 

cent, or more). 

(2) A high current efficiency is obtained (95 per cent.) 

and no oxygen is evolved. 

(3) The purity of the caustic soda is high, 99 per cent. 

with I per cent, chloride and carbonate. 

The chief disadvantages are the high cost of mercury, 
of which about 70 tons are needed for a 6000 H.P. plant, and 
the higher voltage required (4*3 volts) as against 3*8 volts 
for a diaphragm cell. On the other hand it is claimed by 
mercury cell advocates that the voltage required is actually 
less than for the average diaphragm cell. However, the 
voltage seems in both cases to be about 4 volts, and for the 
Townsend diaphragm cell it is 3*4 to 3*6 volts. 

The initial cost of the mercury is, perhaps, not excessive 
when regard is paid to the cost of evaporating the much 
weaker caustic liquors obtained from diaphragm cells. 

The earliest mercury cell was that of H. Y. Castner 1 
(1892). This was a comparatively small cell of the rocking 
type, and its subsequent improvement, in conjunction with 
Kellner, resulted in the Castner- Kellner Cell. The construc- 
tion of the Castner mercury cell is shown in Fig. 40. The 
cell is constructed of slate or earthenware, is rectangular 
in shape, about 4 ft. X 4 ft. X 6 in. deep, and requires about 
200 Ib. of mercury. Two partitions which divide the cell 

1 Eng. Pats., 16046 (1892); 10584 (1893). Chem. Trade Journal* 
1894, 15, 211 ; Electrochem. Ind., 1902, 1, 12. 



into three equal compartments, and reach to wifhin T V in. 
of the bottom of the cell, fit into groves, so that a very 
thin layer of mercury permits communication between the 
compartments. The two outer anode compartments contain 
brine, and there are two graphite anodes AA ; the chlorine 
evolved is drawn ofT by pipes at the top of each anode com- 
partment. Water circulates through the middle compart- 
ment C, and this contains an iron grid cathode. A rocking 
motion is given to the cell by an eccentric wheel W, which, 
as it turns, causes the cell to rise and fall at that end, so that 
the mercury circulates backwards and forwards. The sodium 
which is discharged on the mercury in the anode compart- 

FIG. 40. 

ments is transferred, by the rocking movement of the cell, 
to the cathode compartment where it is attacked by the water 
circulating through C, and converted into caustic soda, while 
an equivalent amount of hydrogen is evolved. 

The mercury in the cathode compartment, having given 
up its sodium, ultimately finds its way back to the anode 
compartment to be recharged with alkali metal. According 
to one account of the working of the cell, 1 144 cells are used, 
each taking 560 amps, at 4 volts, and under these conditions 
each cell gives f gallon of 20 per cent, soda per hour. Forty 
tons of bleaching powder are made per week from the chlorine 
evolved. The mercury is cleaned by mechanical means, and 
by treatment with dilute nitric acid, at intervals. Loss of 

1 Journ. Soc. Chem. Ind., 1913, 32, 995, 


mercury during working is small, and amounts to some- 
thing like 2 per cent, per annum. Care is needed to prevent 
any considerable formation of hydrogen in the anode com- 
partment at the mercury, otherwise explosions will occur. 
The current efficiency of the cell is stated to be 91 per 
cent, and the energy efficiency 52*3 per cent. 

FIG. 41. 

The Castner-Kellner cell is shown in Fig. 41 to consist 
of two compartments separated by a non-porous earthenware 
partition P ending in a metal cap, and by arranging that the 
floor has a slight incline, the cell can be made large since no 
rocking mechanism is needed. The mercury flows out at the 

FIG. 42. 

lower end and is pumped back to the anode compartment 

Concentrated brine is contained in B, and water in 
the cathode compartment A. The secondary cathode S 
diminishes the electro-positive character of the mercury 
so that it does not oxidise easily and remains clean for a 
considerable time. 


Kellner Air-Pressure Cell. A short account of this cell 
is given by Taussig 1 in an article on electrolytic cells. The 
circulation of the mercury in this cell is brought about by the 
aid of compressed air. 

Solvay-Kellner Cell. 2 This is a comparatively large 
stationary cell in which the mercury flows through the cell 
by gravity (see Fig. 42). 

A row of carbon anodes dips into the brine which enters 
at A ; the brine on its passage across the cell is decomposed, 
and the resulting caustic liquor flows out at B. The mercury 
enters at C, and leaves by D after giving up its charge of 
sodium. Chlorine is drawn off by the pipe E. This cell 
is in use at Jemeppe, in Belgium, and at Weston Point, 

The Whiting Cell? In this cell, provision is made for the 
periodic removal of the amalgam. After the mercury has 
acted as cathode for two minutes it is automatically removed 
to another compartment where the sodium is acted upon 
by water, and at the same time its place is taken by fresh 
mercury. The temperature is kept below 40 C., there is 
little risk of chlorate formation, and the carbon anodes are 
found to last well. 

The chlorine evolved contains 2 per cent, of hydrogen. 
Voltage used is 4 volts, and the current efficiency is 90-95 
per cent. The plant at the Oxford Paper Works, Rumford, 
M.E., turned out, in 1910, eight tons of caustic soda per day, 
and the corresponding amount of chlorine was converted into 
bleach. The caustic liquor has an average strength of 20 
per cent. 

Whiting has pointed out that cells which promised well 
when worked on a laboratory scale have often not been 
successful on an industrial scale. He mentions the instances 
of the Rhodin cell and the cells devised by Bell Bros., which 
involved those concerned in great financial loss. He states 
further, that even when a process is intrinsically sound, there 

1 Chem. Zeit., 1909, 33, 588 ; Trans. Faraday Soc., 1909, 5, 258. 

2 D.R.P., 104900 (1898). Trans. Faraday Soc., 1910, 5, 258. 

3 Trans. Amer. Electrochem., 1910, 17, 327. 



are often practical difficulties which cannot always be sur- 
mounted, and he quotes the example of attempts, in Japan, 
to work the purchased rights of the Castner Process for 
caustic soda, which up to 1910 had not proved successful 
and had occasioned heavy financial loss. The chief diffi- 
culty to be overcome in the mercury process is, undoubtedly, 
the proper transfer of the sodium amalgam from anode to 
cathode compartment. On the other hand, the process, in 
the hands of the Castner Alkali Co., at Niagara was a success 
almost from the start. 

Rhodin Cell?- The cell is not now in use. It was worked 
for a short time at Sault Ste. Marie by a company which 

FIG. 43. 

purchased the patent rights from Rhodin, but the venture 
failed as the cell did not come up to expectations when 
employed on a large scale. 

The anodes A of graphite are contained in a hood or 
bell of earthenware (Fig. 43), which is rotated at a speed of 
30 r.p.m. At the bottom of the shallow iron dish (diameter 
5 ft.) is a layer of mercury which forms the cathode C. 

By the rotation, the sodium amalgam, formed under the 
bell, is centrifuged outwards, attacked by water when outside 
the bell, and the sodium converted to caustic. The anode 
compartment contains brine, and the chlorine leaves by a 
pipe B in the top of the bell. On losing its sodium, the 

1 D.R.P., 102774 (1896). Journ. Soc. Chem. 2nd., 1902, 21, 449; 
Zeitsch. Elektrochem., 1903, 9, 366. 



mercury, being denser than the amalgam, falls to the floor of 
the annular space and is driven back to the centre where it 
again takes up more sodium. 

The cell was used by the Canadian Electrochemical Co., 
and was almost the first electro-chemical process to work at 
Sault Ste. Marie. 

The Edser-Wilderman Cell. 1 This cell is made of iron 
lined with patent ebonite which resists the action of both 
chlorine and caustic soda. It is divided into anode and 
cathode compartments by a series of circular V-shaped chan- 
nels, fixed one over the other. This partition or diaphragm 

FIG. 44. 

is made continuous by partly filling each channel with mer- 
cury so that the two compartments only communicate through 
the mercury seal thus arranged (see Fig. 44). The inner or 
anode compartment is filled with brine ; it contains graphite 
anodes AA and a stirring arrangement BB by which the 
mercury in the channels is kept in motion. This motion 
causes the amalgam which is formed to flow to the cathode 
side of the channel, i.e. the caustic soda compartment, 
where it becomes converted into caustic. The stirrer consists 
of ebonite blades with teeth dipping into the channels of 

1 Eng. Pats., 18958 (1898) ; 22902 (1900) ; 9803 (1902). Trans. 
Faraday Soc.^ 1909, 5, 258 ; Trans. Amer. Electrochem., 1912, 22, 445. 


mercury, and pieces of carbon C in electrical connection 
with the mercury in the outer compartment facilitate the 
decomposition of the amalgam. 

In the early form of this cell the inventor depended upon 
the difference of surface tension between sodium amalgam 
and mercury for the transfer of amalgam in the channels, 
but the stirrer was found to greatly increase the value of the 
cell. The strength of caustic soda produced is 20 per cent, 
and does not often contain more than '2 per cent, of chloride. 
A high current density can be used, as much as 60 amps, per 
dm 2 ., since the mercury circulation is very efficient, and very 
little hypochlorite and chlorate are produced. 

The Castner and Solvay cells are required to work with 
a current density of 6-8 amps, per dm 2 , since if higher, solid 
amalgam is formed, but in the Wilderman cell the circulation 
of the mercury is so efficient that there is no risk of forming 
solid amalgam, even with a much greater current density. 
This means a smaller cell for equivalent output, with conse- 
quent saving of floor space. 1 The current efficiency is high, 
90-9-2 per cent. A large installation was in use in 1910, at 
the works of the Zellstoff-fabrik, Waldhof, Mannheim, for 
producing 10,000 tons of bleach per year. 


The use of a bell of earthenware or iron, to enclose the 
anode, is the chief feature of these cells. 

There is no diaphragm, but advantage is taken of the 
motion of the electrolyte, as a whole, to carry the caustic 
liquor away from the anode and prevent its contact with 
chlorine. The bells are usually made of iron and are cemented 
inside to form the anode compartment; alkali chloride enters 
the anode compartment and a distributing arrangement 
ensures the regular distribution of the brine to the anolyte, 
Sometimes the outside of the bell, which is rectangular, forms 
the cathode, at other times an iron cathode surrounds the 


1 Int. Cong. App. Chem.^ 1912, X, 185. 



A sharp alkali boundary, or a neutral layer, forms just 
inside the mouth of the bell, and the migration of OH' ions 
to the anode is counteracted by the steady flow of brine from 
the bell. Caustic soda solution overflows continually from a 
pipe in the upper part of the cathode compartment. 

Sometimes as many as twenty-five bells are hung in a 
cement vessel which forms the cathode chamber ; l each bell 
contains a graphite anode and each is coated with iron sheet 
or gauze to act as cathode. All the bells are connected in 
parallel, and saturated salt solution passes through an opening 
in the anode and is distributed by passing through a number 
of small holes as shown in Fig. 45. The heavy caustic sinks 

FIG. 45. 

to the bottom of the trough and is removed by the overflow 
pipe B. The openings AA, in the bell, are connected so 
that the gas pressure remains constant throughout the bells. 
Chlorine is drawn off through the pipe C. 

Billiter-Leykam Cell?' This is a modified form of the 
Billiter-Siemens diaphragm cell, and it is also an improve- 
ment on the ordinary bell cell. From Fig. 46, it will be seen 
that each cell contains one bell which is an inverted cement 
box A. The distance between anode and cathode can be 
altered so that the resistance between them can be controlled 
and the voltage adjusted. 

The bell contains a chlorine outlet B, and an inlet for 

1 D.R.P., 141187 (1900). 

2 Eng, Pat., 11693(1910). 

Trans. Faraday Soc., 1913, 9, 3. 



brine D ; the graphite anode is cemented into the bell. The 
cathode C is of iron, T-shaped, and is encased in asbestos to 
facilitate the removal of hydrogen. Caustic soda falls to the 
bottom and flows out through the pipe H. The current 
density used is 5-7 amps, per dm 2 , and the pressure 3*1 to 
3'2 volts ; each K.W.H. gives "45 kg. of caustic (NaOH) and 
4 kg. of chlorine. 

Since the direction of the liquid flow does not change at 
the bottom of the bell, the rate of flow is practically constant 
at all points, and larger units can be employed, with higher 
temperatures, than in that type of bell cell where the liquid 
is obliged to turn up, at the edge of the bell, in order to reach 

FIG. 46. 

the outlet. This cell is considered to be one of the most 
efficient non-mercury cells for producing caustic soda. 1 

The Acker Process? 1 This is a process for electrolysing 
fused salt, with carbon anodes and a molten lead cathode. 

The cell is made of cast iron and lined with magnesia 
brick ; its construction is such that the molten lead, charged 
with 20-25 per cent, of sodium, can be skimmed away me- 
chanically to a steam-jet chamber where it is decomposed, 
and the metallic sodium converted into anhydrous caustic 
soda. After the removal of the sodium, the molten lead is 
returned to the electrolysis chamber. Fig. 47 shows how the 

1 Journ. Soc. Chem. Ind., 1913, 32, 995. 

2 Trans. Amer. Electrochem., 1902, 1, 165 ; Zeitsch. Elektrochem.^ 
1903, 9, 364 ; Electrochem. Ind., 1906, 4, 477. 


anodes are arranged so that they dip into a 6-in. layer of 
fused salt and approach to within one inch of the lead cathode 

FIG. 47. 

upon which the fused salt rests. The lead-sodium alloy 
flows continuously to the steam-jet chamber S where steam 
is blown in at a pressure of two or three atmospheres. 

FIG. 48. 

Fig. 48 shows a section through the steam chamber which 
explains how the sodium alloy, carried up at a temperature 


of about 500 C, falls over into the caustic vessel and 
deposits any molten lead. Caustic soda flows by the spout to 
a receiving vessel. The chlorine gas contains some oxygen 
and is converted into bleach in a plant adapted for use with 
dilute chlorine. Anode about 300 amps, per dm 2 , and 
the working temperature about 850 C. Current efficiency 
is usually 93-94 per cent, and each cell gives approximately 
580 Ib. of solid caustic per day, containing 98 per cent. 
NaOH. One ton of caustic requires about 5000 K.W.H., 
hence one K.W. year gives about 1*4 tons. The cell was 
used with success, at Niagara, between 1900 and 1907. 

The Vautin Cell?- Fused salt is used in this process, 
other chlorides being added to lower the melting point. 
Carbon anodes are immersed in the fused electrolyte, and the 
cathode is of molten tin which is withdrawn at intervals 
when sufficiently rich in sodium, by an outlet pipe at the 
bottom of the cell. 

Caustic soda production by electrolysis of fused chloride, 
using a molten lead cathode, is the subject of German Patent 
189474 (1906). 

From the foregoing description of the various alkali- 
chlorine cells, it is evident that electrolytic production is 
a widely established industry. Almost every European 
country has works of this class, and in America many 
small plants are working in connection with wood pulp 

The manufacture of caustic soda, bleaching powder, 
chlorates and perchlorates is certain of rapid future develop- 
ment, and the production of various chlorinated organic 
bodies will serve to utilise the chlorine produced. The 
Leblanc process works will probably find their strongest line 
of development to be sulphur products. In short, the alkali 
works of the future will centre around either chlorine or 

The ammonia-soda process stands alone, but should the 
chloride waste liquors become utilised, as for example in the 

1 Eng. Pat., 13568 (1893) ; 9878 (1894). Journ. Soc. Ckem. Ind., 1894, 
13, 448. 


production of electrolytic zinc, it will become more closely 
allied to the chlorine alkali industry. 

For every ton of caustic soda produced by electrolysis, 
there is evolved sufficient chlorine to form 2\ tons of bleach- 
ing powder. Hitherto, the chlorine has been difficult to 
dispose of and the electrolytic soda industry will progress 
more steadily if a good market for this gas is assured. 


The Purification of Electrolytic Chlorine, U.S. Pat., 1166524. 

The Electrolytic Soda Industry, M. L. Moynot, Monit. Scient., 1907, 
66, 586. 

Electrolytic Production of Caustic Soda, C. P. Townsend, Electro- 
chem. Ind.) 1902, 1, 23. 

Electrochemistry at Sault Ste. Marie, J. W. Richards, Electrochem. 
Ind., 1902, 1, 86. 

Influence of some Impurities in Salt on the Yield of Caustic Soda by 
the Mercury Process, Walker and Paterson, Trans. Amer. Electrochem., 
1903, 3, 185. 

Formation of Metallic Dust at Cathodes during Electrolysis of Alkali 
Chloride Solutions, F. Haber, Trans. Amer. Electrochem., 1902, 2, 189. 

The Cost of Alkali Chloride Electrolysis, Engelhardt, Met. andChem. 
Eng., 1911,9, 489. 

Present Position and Future Prospects of Electrolytic Alkali, J. B. 
Kershaw, Trans. Faraday Soc., 1907, 3, 38. 



WHEN an aqueous solution of alkali chloride is electro- 
lysed in a cell without a diaphragm, the chlorine liberated at 
the anode reacts with the hydroxide which is formed at the 
cathode, and the following reactions take place according to 
the amount of chlorine or soda present. 

(i) 2NaOH + C1 2 = NaCIO + NaCl + H 2 O. 
(a) NaOH-f C1 2 =NaCl + HClO. 

The hypochlorite formed commences to decompose as its 
concentration increases, and the rate of decomposition rises 
with the concentration. 

Decomposition of the hypochlorite is due partly to the 
reducing action of cathodic hydrogen, and partly to the 
action of water on the hypochlorite ion at the anode 

NaCIO + H 2 =H a O + NaCl, 

6C10' + 3H 2 = 2HC10 3 + 4C1' + 4H* + 30. 

The hypochlorite is further decomposed by the action of 
hypochlorous acid 1 in the following manner 

2HC1O + NaCIO = NaClO 3 + 2HC1. 

Although a weakly alkaline solution of hypochlorite is 
stable, and likewise a weakly acid solution of hypochlorous 
acid, when mixed, they react as shown by the last equation. 

Chlorate is also formed by the spontaneous decomposition 
of the hypochlorite, as well as by its anodic oxidation 

3 NaC10 = NaC10 3 + 2NaCl, 
NaCIO + 20 = NaC10 3 . 

1 Journ. pr. Chern., 1899, 59, 53; 1901, 63, 141. 


Both chlorine and oxygen may be liberated at the anode 
but the oxygen discharge, at platinum electrodes, is subject 
to a high overvoltage (about i volt), and chlorine is therefore 
more readily discharged from a dilute solution. The voltage 
required for the continuous discharge of oxygen from platinum 
is r8 volts, and for the discharge of chlorine it is 1*37 volts, 
so that, under ordinary conditions, hydrogen is liberated at 
the cathode, and near to it the electrolyte becomes alkaline, 
while an equivalent amount of chlorine is discharged at the 

Knowledge concerning the reactions which take place 
during the electrolysis of alkali chloride solutions is largely 
due to Foerster and his co-workers. The following facts 
have been ascertained. 

The interaction of chlorine with hydroxyl, to produce 
hypochlorous acid, is fairly complete, C1 2 + OH' == HC1O + 
CT, but in solutions which are more concentrated it is possible 
to detect free caustic soda and chlorine. Bromine and iodine 
are much less reactive, and in these two cases a considerable 
quantity of free halogen can exist in alkaline solution. 

Hypochlorous acid forms water with hydroxyl, especially 
if the acid solution be moderately concentrated : HC1O + 
OH' = H 2 O + CIO'. 

Combining the last two equations we get, C1 2 + 2OH' = 
H 2 O + CIO' + Cr, the ionic expression of the well-known 
equation C1 2 + 2NaOH = H 2 O + NaCIO + NaCl, which 
represents, fairly well, what obtains in practice. The most 
stable system resulting from the reaction of chlorine upon 
alkali is NaCl + O, and not NaCIO, owing to the instability 
of the hypochlorite, 2C1O' = 2d' + O 2 , and the oxygen 
liberated helps to form chlorate thus 

2HC10 + CIO' = ClO's + 2H' + 2C1'. (i) 

If only one molecule of chlorine reacts with two molecules 
of NaOH, the rate of chlorate formation is very slow, but the 
formation of additional hypochlorous acid accelerates chlorate 
production. In practice this result is secured by allowing 
slightly more chlorine to be discharged than is required by 


the equation, C1 2 + 2NaOH = H 2 O + NaCl + NaCIO, as a 
result, hypochlorous acid is formed, and since there is not 
sufficient alkali to neutralise this, the reaction denoted by 
equation (i) can proceed. 

Chlorate production takes place at the anode where hypo- 
chlorous acid accumulates, but its formation is suppressed by 
maintaining a low temperature. It is possible to prevent the 
cathode reduction of hypochlorite by adding alkali chromate x 
to the bath ; this possibly acts favourably by the formation of 
a layer of chromium oxide on the cathode, which is affected 
by the discharged hydrogen before the latter has an oppor- 
tunity of attacking the hypochlorite. Addition of other 
compounds has a beneficial result ; such are, certain aromatic 
sulphur compounds and some calcium salts. 2 

Concentration of hypochlorite is limited by the discharge 
of hypochlorite ions ; it is also attacked by water, as indi- 
cated in an earlier equation, a change which is represented 
by Foerster and Muller 3 by the equation 

6C10' + 3H 2 = 6H' + 2C10' 3 + 4C1' + 3<3 + 60. 

Platinised platinum anodes suppress the discharge of 
hypochlorite ions and the loss due to chlorate formation ; 
and a high chloride concentration favours Cl' discharge which 
hinders CIO' discharge and so favours concentration of 

The chief conditions to be observed in order to obtain a 
high yield of hypochlorite are 

(1) Concentrated chloride solution which also has the 

effect of increasing conductivity. 

(2) Low temperature and high anode current density. 

(3) Presence of alkali chromate in the bath and the use 
of platinised platinum anodes. 

(4) Magnesium and calcium salts, if present in the salt 
used, should be removed, because Mg(OH) 2 and 
Ca(OH) 2 are precipitated during electrolysis, the 

1 Zeitsch. Elektrochem.) 1899, 5, 469; 1901, 7, 398; 1902, 8, 909. 

2 D.R.P., 141372 (1903); 205087 (1908). 

3 Zeitsch. Elektrochem., 1902, 8, 665. 


solution consequently becomes acid and chlorate 
formation is increased. 

It is not always profitable to adhere strictly to these con- 
ditions ; it is advisable to make a dilute hypochlorite solution 
because the energy efficiency falls off with increase in hypo- 
chlorite concentration. 

Technical Cells. Bipolar electrodes are generally used 
because compact plant can be made and the number of 
exposed metallic connections is reduced ; there is, however, 
some danger of shunt current losses which are greater with a 



FIG. 49. 


high concentration of salt, since the drop in voltage between 
the electrodes is as high as 4-6 volts. 

Kellner Cell} The vertical type of this cell is widely used 
and is supplied with several electrolytic bleaching plants 
which are on the market, notably, those of Siemens Bros. & 
Co., Mather & Platt, and Siemens & Halske. The cell con- 
sists of a deep rectangular trough of earthenware with sup- 
ports A A (Fig. 49), divided into a number of vertical chambers 
by glass plates which are wound round with platinum iridium 
wire or covered with netting of the same material ; these 
plates form the bipolar electrodes. The end electrodes are 
provided with platinum coated leads. 

1 D.R.P., 99880 (1894) ; 104442 (1896). 



Brine enters the bottom of the cell at B and overflows at 
the top through slits or spouts OO into the main supply 
tank underneath. The liquor in this tank is cooled and 
returned to the electrolyser, this circulation being continued 
until the hypochlorite has reached the required concentration. 
Generally, there are twenty compartments in each unit, taking 

FIG. 50. 

a pressure of no volts and a current density of 50-100 amps, 
per dm 2 . The brine used contains 10 per cent, of chloride 
with -i to -5 per cent, of sodium chromate. 

The first electrolytic bleach cell used on the large scale 
was that of Hermite which was employed in Scotland and in 
France about 1888. The pioneers in the electrolytic bleach 


FIG. 51. 

process were Hermite in England, Karl Kellner in Vienna, 
and Stepanoff in Russia, whose process has been working 
since 1891. 

In another form of the Kellner cell, horizontal electrodes 
are used. 1 Each cell is again divided into a number of com- 
partments by vertical partitions of glass, and these compart- 
ments communicate by channels alternately on each side (see 
1 D.R.P., 165486 (1903). 


Figs. 50 and 51). The bipolar electrodes are o'f platinum, 
each sheet being bent under the partitions so that one half 
forms the cathode of one compartment, and the other half 
forms the anode in the next compartment. 

The cooled electrolyte circulates through the cell by 
gravity, its concentration is .10-15 P er cent, of sodium 
chloride with '5 per cent, of alkali chromate. The anodes 
are very close to the bottom of the cell and the cathodes 
where gas is evolved are 5 mm. above. There are thirty-six 
compartments in a 220 volt unit, which takes 60 amps, at a 
temperature of 21 C. 

The Schuckert Cell}- This cell is used by the Siemens- 
Schuckert Co., Berlin. Bipolar 

electrodes of platinum-iridium -f 

are used, and each earthenware 
cell is divided into nine com- 
partments and contains a cool- 
ing coil. Two cells are generally 
connected in series and take a 
pressure of no volts at the ter- 
minal electrodes. Efficient cir- 
culation is assured by causing 
the electrolyte to pursue a zigzag 

course on its way through the cell. Calcium chloride and 
sodium rosinate are dissolved in the electrolyte. 

Haas-Oettel Cell? This cell is largely due to the efforts 
of Professor Oettel of Zurich, whose ideas show considerable 
novelty and ingenuity. He invented a cell in which platinum 
electrodes are replaced by carbon or metal, and in which 
automatic circulation is possible. According to Reuss 3 this 
cell is likely to supplant all the older hypochlorite cells. The 
cell consists of two earthenware vessels, one within the other, 
and the brim of the smaller vessel is just below the surface 
of the brine which fills the larger outer vessel (Fig. 52). The 

1 D.R.P., 141724 (1902). Electrochem. Ind., 1903, 1, 439- 

2 D.R.P., 101296(1896); 114739(1900). Zeitsch. Elektrochem., 1901, 

7, 3i9- 

3 Journ. Soc. Dyers and Colorists^ 1911, 27, no. 









FIG. 52. 


electrodes, of metal wire or carbon, are bipolar, and are 
insulated at top and bottom by glass insulating caps ; these 
caps prevent the carbon from wearing at the top edges, and, 
at the bottom of the trough, they prevent the sediment from 
producing a short circuit. 

Openings between the electrodes in the bottom of the 
small trough allow cold brine to enter the electrode cell, 
where the increase in temperature and evolution of gas cause 
the liquor to rise. The bleach liquor therefore rises to the top 
and flows off, while fresh brine takes its place from below. 
Earthenware channels are provided on both sides, at the top, 
for the overflow of bleach liquor. 






FIG. 53. 

FIG. 54. 

Schoop Cell}- In this cell, shown in section (Fig. 53) and 
in plan (Fig. 54), the brine flows through a number of chan- 
nels from cell to cell in cascade fashion. The electrodes of 
platinum foil project into the channels, and they are arranged 
in such a manner as to cause the intermediate sheets to 
become electrodes by induction (see Fig. 54). 

Hermite Process at Poplar? Electrolytic bleaching liquors 
are made by the Poplar Borough Council, for disinfecting 
purposes, by the Hermite process. The power consumption 
is about 7-2 K.W.H. per kg. of active chlorine, and for this 
quantity 13*5 kgs. of salt are necessary. The concentration 

1 D.R.P., 118450, 121525 (1899). 

2 Trans. Faraday Soc. } 1906, 2, 182. 


of the salt solution is 5 per cent, and it contains, in addition, 
I per cent, of MgCl 2 ; the working temperature is 30 C. 

Most hypochlorite cells take 6-7 K.W.H. per kg. of 
active chlorine, and this only represents an efficiency of 20-25 
per cent. The losses are due mainly to low current efficiency 
and the high overvoltages at the electrodes. 


The formation of chlorate by electrolysis is dependent 
upon the reaction 2HC1O + NaCIO = NaClO 3 + 2HC1, 
which is accelerated by slight acidity and by rise of 
temperature to 60 or 70 C. 

The solution used contains about 25 per cent, of 
alkali chloride, and the addition of alkali chromate is 

Foerster and Muller 1 have thoroughly examined the con- 
ditions favourable for chlorate formation, and shown the 
necessity of a slight acidity, such as may be produced by CO 2 , 
in order to liberate the hypochlorous acid, necessary for the 
above reaction. They proved the necessity of preventing 
cathodic reduction either by employing a high cathode I.D. 
or by adding alkali chromate, and the need for a temperature 
above 40 to prevent formation of perchlorate. 

The earliest technical cell was devised by H, Gall and 
de Montlaur 2 and was used in Switzerland for a number of 
years, even as late as 1900. A diaphragm was employed, 
the anodes were of platinum-iridium, and the cathodes of 
platinum or nickel. The solution contained 25 per cent, of 
chloride, at a temperature of 45-50 C., and the alkali hydrate, 
from the cathode compartment, circulated through the anode 
compartment and reacted with the chlorine there evolved to 
form chlorate. Current density was 50 amps, per dm 2 , and 
each cell took 4-5 to 5 volts. 

Other cells of the diaphragm type were invented by 
Hurter in 1893, and by Blumenberg 3 in 1894. 

1 Zeitsch. anorg. Chem., 1899, 21, i, 39. 

2 Eng. Pat., 4686(1887). 

8 Eng. Pats., 15396 (1893) ; 9129 (1894). 

I2 4 


In 1899, Imhoff 1 patented the addition of alkali chromate 
to the bath, and this rendered the diaphragm unnecessary. 

The process of Gibbs 2 for chlorate production is used by 
the National Electrolytic Co. at Niagara. 

Each cell consists of a wooden trough lined with lead 
and divided into four parts. The anodes B are of sheet lead 

FIG. 55. 

covered with platinum foil, and the cathodes C consist of 
copper wires fixed vertically in the cell by insulating bars 
O (Fig. 55). There is no diaphragm, the temperature is 
maintained at 60-70 C., and the brine flows through the 
cell at a rate of 28 litres per hour ; anode I.D. is 500 amps, 
per ft 2 . This particular plant was capable of producing 

1 U.S. Pat., 627063 (1899). 

2 Electrochem. Ind., 1902, 1, n. 


4000 lb. of chlorate per day. Alkali chloride is fed in 
through G, and the chlorate liquor drawn off through H. 

In continental factories the process devised by Lederlin 
and Corbin x is widely used, and is the subject of numerous 
patents. The two sections of Fig. 56 show a cement tank 
through which the brine (25 per cent.) circulates, at a 
temperature of 70 C. The platinum bipolar electrodes A 
are set close together in ebonite frames D, the distance 
between two electrodes being 1*5 cms. The frames are kept 
in position by wood supports B C. The chloride enters the 
tank by the pipe E, and after circulation the chlorate 
formed leaves at F. 

Chlorides of magnesium or calcium are added to the 

FIG. 56. 

bath, as their presence is found to assist chlorate formation, 
and one gram of potassium chromate is added per litre. The 
anode I.D. is 10-20 amps, per dm 2 . (90-180 amps, per ft 2 .), 
and the voltage between adjacent electrodes is 4-5 to 5 
volts. If potassium chlorate is being made, the liquors are 
drawn off when sufficiently strong, and the chlorate allowed 
to crystallise. In making sodium chlorate, when the con- 
centration has reached 500-600 gms. of NaClO 3 , and 100 
gms. of sodium chloride per litre, the liquor is evaporated, to 
remove crystals of chloride, and then, on further cooling or 
evaporating, the chlorate separates out. Provided the con- 
centration of the chloride is maintained sufficiently high there 
is no fear of perchlorate formation. 

Substances other than chromate are added to the electro- 

1 Fr. Pats., 226257 (1892); 238612 (1894); 110505 (1898); 136678 
(1901); 283737(1904)- 


lyte by various makers to prevent cathodic reduction, namely, 
aluminium salts, clay, or silicic acid by the United Alkali 
Company ; 1 alkali fluorides by Siemens and Halske ; 2 and 
vanadium compounds by the Solvay Works. 3 In the process 
of Threlfall and Wilson, 4 free chlorine is formed by making 
the current density greater at the anode than at the cathode. 
A process patented by A. G. Betts has been devised for 
obtaining potassium chlorate by electrolysis of the chloride 
solution which results from the suitable treatment of felspar. 
The resulting mixture of sodium and potassium chloride is 
electrolysed with carbon anodes, and magnesium cathodes 
are employed to avoid the expense of platinum. 5 


The formation of perchlorate is dependent upon the 
prior formation of chlorate, and it is usual to employ a bath 
of chlorate, and to convert this, at a low temperature, into 
perchlorate by anodic oxidation ; NaClO 3 + 2 OH' + 20 = 
NaC10 4 + H 2 O. 

During the electrolysis of sodium chloride solution, the 
chlorate formed will be partly transformed into perchlorate 
if the concentration of the chloride is low and the current 
density small. Above a concentration of 10 per cent, there 
is no formation of perchlorate to speak of, provided the 
solution is warm or hot, so that, conditions favouring per- 
chlorate formation from chloride are : a dilute solution of the 
latter (under 10 per cent.), and low temperature accompanied 
by low current density. 

In order to convert chlorate to perchlorate the following 
conditions have been laid down by Winteler 6 

(1) Low temperature at anode (artificial cooling). 

(2) Acid solution at anode. 

(3) Current density 4-12 amps, per dm 2 . 

(4) High concentration of chlorate (60-70 per cent.). 

1 Eng. Pat, 1017 (1899). 2 D.R.P., 153859(1903). 

3 D.R.P., 174128 (1905). 4 D.R.P., 143347 (1902). 

5 U.S. Pat., 918650 (1909). 6 Chem. Zeit., 1898, 22, 89 


According to Couleru, 1 starting with brine, the elec- 
trolysis is carried on until the concentration of the chlorate 
is about 750 gms. per litre ; the chlorate is then allowed 
to deposit and is afterwards re-dissolved in water and elec- 
trolysed with smooth platinum anodes and iron cathodes. 
Any alkali formed during electrolysis is neutralised, and the 
temperature is kept at 8-10 C. A temperature above 25 C. 
is injurious, and it is necessary as a rule to cool the electrodes. 
Couleru recommends the addition of calcium chloride or 
sodium chromate for obtaining a good yield of chlorate, and 
then a concentrated solution of this salt for the successful 
production of perchlorate. 

The anode I.D. is about 8 amps, per dm 2 , with a pressure 
of 6'5~7 volts. At a smooth platinum anode, the oxygen 
overvoltage is high, and discharge of oxygen is therefore 
retarded. One kg. of perchlorate requires about 3*5 K.W.H. 
Since sodium perchlorate is very soluble and hygroscopic, 
it is usually not isolated, but converted into the corresponding 
potassium or ammonium salt. 


The chief sources of bromine are the salt deposits at 
Stassfiirt in Germany, and a few less well-known deposits in 
the United States. In addition to the usual method of pre- 
paring bromine by the chlorination of concentrated bromide 
mother-liquors, a certain amount is obtained by electrolysis of 
similar liquors containing a mixture of chloride and bromide. 

Bromine is hydrolysed like chlorine on discharge, but 
not to the same extent, Br 2 + H 2 O = HBrO + H* + Br'. 

In the case of iodine there is no perceptible hydrolysis. 
The action of hydroxyl upon bromine in the neighbourhood of 
the anode may be thus expressed : Br 2 + OH' = HBrO + Br', 
and the hydroxyl also acts upon the hypobromous acid 
thus: HBrO + OH' = BrO' + H 2 O, while the hypobromite 
ion reacts with hypobromous acid to form bromate, thus 

2HBrO + BrO' = BrO' 3 + 2H" + 2Br'. 
1 Chem. Zeit., 1906, 30, 213. 


Bromate formation, however, takes place with a velocity 
which is one hundred times as great as that of chlorate 
formation. 1 

The ratio of bromine to chlorine in the liquors used lies 

between - and - - ; but if the current density be low 

there is no risk of chlorine discharge. 

The Kossuth Cell for bromine consists of a long open 
cement trough containing a number of vertical carbon plate 
bipolar electrodes, which rest on the bottom of the trough 
and rise some distance above the level of the liquid. The 
electrolyte is compelled to take a zigzag course so that it 
passes over the surface of each electrode on its way through 
the cell. The end plates are connected with the source of 
current and the intermediate plates become electrodes by 
induction (bipolar). 

The working temperature is 60 C., and the liquor 
becomes charged with bromine, which is volatile, while the 
equivalent of magnesium is precipitated in the form of 
hydroxide as the liquor traverses the cell. Each cell con- 
tains thirty electrodes, and takes about 100 amps, at a 
pressure of 100 volts. The amount of energy needed for 
producing one kg. of bromine is about 3 K.W.H. 

The Dow Cell employed in America is a diaphragm cell, 
used apparently because the concentration of the chloride is 
rather high, and without a diaphragm considerable loss of 
bromine occurs by conversion into bromate by the chlorine 
which is discharged. 


These are obtained by electrolysis of concentrated bromide 
solutions, at 40-50 C., containing a small amount of alkali 
bichromate ; this latter produces a slight acidity which acceler- 
ates the reaction 2HBrO + BrO' = BrO' 3 + 2tT + 2Br', and 
the chromate also prevents the cathodic reduction of bromate 
and hypobromite. 

1 Zeilsch. Elektrochem., 1897, 3, 474 ; 1904, 10, 802 ; 1905, 11, 57 ; 
1910, 16, 321 ; Chem., 1901, 63, 141. 


Smooth platinum anodes and graphite cathodes are gener- 
ally used, and I.D. at the anode is 10-15 amps, per dm 2 . 


A patent was granted to Parker & Robinson, 1 in 1888, 
for the production of iodine by the electrolysis of alkali 
iodide containing a soluble sulphate and some free acid. 
The anode and cathode compartments were separated by a 
porous diaphragm, and carbon or platinum anodes with iron 
cathodes were used. 

The cathode compartment was filled with caustic soda, 
and the anode compartment contained the iodide solution. 
The iodine which separated in the anode compartment was 
washed and dried. 

In a process patented by B. Rinck 2 for the production of 
bromine and iodine, the electrolyte is a solution of alkali 
halide, and a diaphragm of asbestos is used. The anode 
compartment contains concentrated brine in which carbon 
plate anodes are immersed. The cathodes are of iron, and 
are immersed in the stream of electrolyte which slowly 
moves through the cell. The iodine or bromine is discharged 
at the anode and dissolves in the brine, from which it can 
be liberated when a sufficient concentration has been reached. 


Some Factors in the Cost of Hypochlorite Manufacture, Int. Cong, 
of Applied Chem., 1912, Xa, 127. 

Cells for the Production of Electrolytic Bleach, W. H. Walker, 
Electrochem. Ind., 1902, 1, 439. 

Elektrochemie Wasseriger Losungen, Fritz Foerster, 1915, 2nd edition. 

1 Eng. Pat., 11479 (1888) 

2 D.R.P., 182298 (1906). 


CERTAIN substances can be produced by electrolysing a 
suitable alkali salt solution with an attackable anode. For 
example, if a solution of sodium sulphate be electrolysed, 
using a copper anode, the copper is dissolved, forming copper 
sulphate which in the neighbourhood of the cathode, where 
sodium hydroxide is formed, is converted into copper hydr- 
oxide. The copper is dissolved at the anode, and sodium 
hydroxide in equivalent quantity is formed at the cathode. 1 

If the substance formed by solution of the attackable 
anode is insoluble, e.g. PbSO 4 , it collects on the bottom of 
the bath beneath the anode. Many patents have been issued 
or processes to manufacture the above-mentioned substances, 
and others of like nature. Most attention has been directed 
to the manufacture of white lead, and a paper by Burgess 
and Hambuechen should be consulted for a full account of 
the literature and patents devoted to this subject. 2 One of 
the chief difficulties met with in the preparation of white lead, 
was the incrusting of the anode with the precipitate there 
produced, which quickly stopped further action. Luckow 
showed that,. in the case of lead sulphate, if a considerable 
quantity of a secondary salt such as sodium chlorate is used 
with the sodium sulphate, then the resulting sulphate of lead 
is formed a very short distance from the anode, and falls 
continuously to the bottom of the cell. 3 White lead is 

1 Zeitsch. anorg. Chem., 1896, 12, 436. 

2 Trans. Amer. Electrochem., 1903, 3, 299. 

3 Zeitsch. Elektrochem , 1903, 9, 797. D.R.P., 91707 (1897) ; 105143 



precipitated continuously if a large excess of chlorate be 
used with the sodium carbonate. 

The chlorate is termed the secondary salt and the carbon- 
ate is the primary salt. The mechanism of the reaction may 
be explained by assuming that the CO" 3 ions are " crowded 
out" by the C1O' 3 ions, so that the lead ions are able to 
travel a short distance towards the cathode before they are 
precipitated by contact with the CO" 3 ions. 

For the production of white lead, Luckow recommends 
a 1*5 per cent, solution composed of 9 parts chlorate and 
I part carbonate of soda, and carbon dioxide is passed in 
during electrolysis. Anode I.D. is '25 amp. per dm 2 , at a 
pressure of 1*4 volts, and it is possible to produce 3*5-4 
kgs. of white lead per K.W.H. 1 

If sodium chromate be used instead of carbonate, then 
chrome yellow is precipitated, and chromic acid must be 
added continuously to replace that which is used up. 

For the production of lead peroxide, the electrolyte is a 
dilute solution of 99*5 parts of Na 2 SO 4 with *5 part of 
chlorate, and the bath is made slightly acid with sulphuric acid. 

Cuprous oxide can be prepared by using a common salt 
solution as electrolyte containing a small quantity of alkali. 
The colour and uniformity of the product are improved by 
the addition of a little sodium nitrate to the bath. 2 

It has been shown that it is necessary to use dilute solutions 
of the precipitating salt, and a low current density in order 
to obtain regular precipitates and prevent the formation of 
crusts. Opposite conditions, however, high concentration 
of precipitating salt and high current density, give finely 
precipitated particles, so that these two opposing sets of 
conditions must be carefully regulated to produce a good 
white lead. 3 

The electrolytic production of zinc white has been 
attempted with some success. 4 

1 Trans. Amer. Electrochem., 1904, 5, 230. U.S. Pat., 644779 ( I 9)- 

2 Eng. Pat., 14310 (1915). 

3 Zeitsch. Elektrochem., 1902, 8, 255; 1903, 9, 275; Journ. Phys. 
Chem., 1909, 256, 332. 

4 Int. Cong. App. Chem., 1909, Sect. X, 45. 

K 2 


Per carbonates. There is, no doubt, a good future for the 
electrolytic production of the salts of the per-acids. 

Potassium percarbonate was first prepared by Constam 
and Hansen. 1 A saturated solution of the carbonate is cooled 
to 10 C, and electrolysed in a divided cell with platinum 
anodes. On a small scale, a porous pot forms the cathode 
compartment, and coiled round the outside is a platinum wire 
anode. Current density used is 30-60 amps, per dm 2 ., and 
during the electrolysis carbonate must be replenished to 
replace that converted into percarbonate. 

Apparently, the carbonate in solution dissociates into the 
ions K* and KCO' 3 , and the reaction at the anode is represented 
by the following equation 

2KCO' 3 + 20 = K a C 2 0, 

Hydroxylamine (NH 2 OH). Tafel showed, in 1903, that 
nitric acid may be reduced electrolytically to hydroxylamine, 
and a reasonable yield of the product obtained. 2 About this 
time several patents 3 were taken out for the manufacture of 
hydroxylamine by electrolysis. Tafel showed that it is pos- 
sible to use a 50 per cent, solution of H 2 SO 4 or a 25, per cent, 
solution of hydrochloric acid with a lead cathode, the nitric 
acid being dropped in continually at a temperature not 
above 20. A cathode of spongy tin was also found to work 
well, and the current density usecl is about '24 amp. per cm 2 ., 
but by stirring the liquor it is possible to raise the I.D. to 
*6 amp. per cm. 

HNO 3 -f- 3H 2 = NH 2 OH + 2H 2 O. 

According to the French Patent 322943 (1903), an anode 
of platinum is used with a tin cathode. Sodium nitrate is 
dropped into the cathode compartment, and the anolyte is 
sodium chloride solution. The yield of hydroxylamine is 
said to be 60-80 per cent., and chlorine is produced at the 
same time. 

1 Zeitsch. Elektrochem.) 1896, 2, 137 ; 1897, 3, 445. 

2 Zeitsch. anorg. Chem., 1902, 31, 289. 

3 D.R.P., 133457, 137697 (1902) ; U.S. Pat., 727025 (1903). 


French Patent 318978 (1903) also relates to the production 
of hydroxylamine by this method. 

Hydro sulphites (Hyposulphites]. Sodium hydrosulphite is 
now made in large quantity by the electrolytic reduction of 
sodium bisulphite. The work of Jellinek has made clear the 
necessary conditions for obtaining a good yield. 1 He showed 
that the thiosulphate formed during the electrolysis is not 
due to reduction but to the decomposition of the hydrosulphite, 

2Na 2 S 2 O 4 + H 2 O = Na 2 S 2 O 3 + 2NaHSO 3 . 

He therefore used a high current concentration, so that 
the current was large compared with the volume of electro- 
lyte, and the hydrosulphite formation then took place at 
a much greater rate than its decomposition. 

The older method of making hydrosulphite by zinc reduc- 
tion of sodium bisulphite is no doubt electrolytic in principle, 
reduction, taking place before hydrogen discharge owing to 
the high hydrogen overvoltage at zinc. 

Zn + 2H* + 2HSO / 3 >Zn" + S 2 O" 4 + 2H 2 O. 

Jellinek used 5N.NaHSO 3 , and obtained a 10 per cent, 
solution of Na 2 S 2 O 4 , using 5 amps, for every 100 c.c. of catho- 
lyte. A low temperature is necessary, and hydrogen must 
not be allowed to leave the anode for the cathode, or the 
hydroxyl will react thus 

HSO' 3 + OH' = HSO' 4 + H- -f 20. 

According to the German Patents 276058, 276059 (1912), 
dilute sodium bisulphite solution is electrolysed in the cathode 
compartment. A neutral salt may be added with good 
results, such as chloride or sulphate, but not a sulphite. 
Temperature is 0-5 C., and during electrolysis, sulphurous 
acid is added to the bisulphite solution. 

According to D.R.P. 278588 (1912), the process is made 
continuous by circulating the bisulphite from a reservoir 
through the electrolytic vessels and back again. When the 
liquor becomes sufficiently concentrated, sodium hydrosulphite 

1 Zeitsch. Elektrochem., 1911, 17, 157, 245. 


separates but vigorous circulation from anode to cathode 
through the diaphragm is necessary. 

According to the French Patent 467443 (1914), hydro- 
sulphite is produced by means of electrolytic zinc sponge. 

Persulphuric Acid and Hydrogen Peroxide. The electro- 
lysis of sulphuric acid (A = 1-5) at a low temperature, with 
high anode current density, yields persulphuric acid. 

This process can be adapted to the formation of per- 
sulphates or hydrogen peroxide, as the latter substance is 
formed by hydrolysis of the persulphuric acid first formed. 

Hydrogen peroxide is produced by the Consortium f. 
Elektrochemische Industrie, Nuremburg, 1 in this way, the 
persulphuric acid first formed being subsequently distilled 
under reduced pressure. 

The process of Pietzsch and Adolp 2 involves the produc- 
tion of potassium persulphate by electrolysis of acid potassium 
sulphate, and the persulphate is subsequently distilled with 
sulphuric acid (A = 1-4). A solution containing 20-30 per 
cent, of hydrogen peroxide is obtained. 

No diaphragm is necessary for the production of per- 
sulphuric acid, but platinum electrodes should be used, and 
a high current density employed in a well-cooled solution. 
The addition of a small quantity of HC1 or HF is beneficial. 3 
The resulting solution contains about 40 per cent, of 
persulphuric acid. 

The production of persulphate has been studied by 
Elbs and Schonherr. 4 In a diaphragm cell a concentrated 
solution of ammonium or potassium sulphate is used in the 
anode compartment, and a moderately strong solution of 
sulphuric acid in the cathode part. It is preferable to omit 
the diaphragm and to use a small amount of alkali chromate 
(about '2 per cent.) in the bath to prevent cathodic reduction. 

The voltage required is about 7 volts, and I kg. of per- 
sulphate is given by 2-4 K.W.H. 

1 D.R.P., 199958, 217538, 217539 (1905). 
* Eng.'Pats., 23158, 23660 (1910). 

3 Zeitsch. Elektrochem., 1895, 1, 41 7. 

4 Zeitsch. Elektrochem., 1896, 2, 245. 


Instead of acidifying the normal salt, the acid salt, KHSO 4 
may be used, and the reaction is facilitated by the presence of 
chlorine and fluorine ions. The current density at the anode 
should be about 50 amps, per dm 2 , and the electrode should 
be of smooth platinum. 

The hydrolysis of persulphuric acid takes place in two 
stages, the intermediate product being Caro's acid 

(1) H 2 S 2 8 + H 2 = H 2 S0 5 + H 2 S0 4 . 

(2) H 2 S0 5 + H 2 = H 2 S0 4 + H 2 2 . 

According to U.S. Patent, 1195560 (1915), hydrogen per- 
oxide is made from ammonium persulphate which is formed 
by electrolysis and then heated under pressure to hydrolyse 
the persulphate. 

(NH 4 ) 2 S 2 8 + 2H 2 = (NH 4 ) 2 S0 4 + H 2 SO 4 + H 2 O 2 . 

The hydrolysed solution is then distilled under reduced 
pressure in an inert gas to separate the hydrogen peroxide. 
The ammonium sulphate is used again for the production of 

Potassium Permanganate. The oxidation of the man- 
ganate melt (formed by fusion of MnO 2 , potash and potassium 
chlorate) is now often effected by electrolysis. 

In the older methods chlorine or carbonic acid is used to 
bring about the following changes 

2K 2 Mn0 4 + C1 2 = 2KMn0 4 + 2KC1. 
3K 2 Mn0 4 + 2C0 2 = 2KMnO 4 -f MnO 2 + 2K 2 CO 3 . 

By electrolytic oxidation the following change is brought 

2K 2 MnO 4 + O + H 2 O = 2KMnO 4 -f 2KOH. 

The advantage of this process is that no manganese 
dioxide is formed, and the potash produced can be used 
again in the fusion process. 

The first industrial cell was devised by Schering, 1 who used 
a diaphragm of cement to separate the two compartments. 

The cell of the Salzbergwerke neu Stassfiirt 2 is shown in 

1 D.R.P., 28782 (1884). 2 D.R.P., 101710 (1898). 

3 6 


section in Fig. 57. The bath contains a solution of potassium 
manganate (K 2 MnO 4 ) which forms the anodic liquor, and 
which is replenished by the gradual solution of fused product 
placed in the metal baskets P. The cathodes are contained 
in porous compartments of cement, and the iron anodes dip 
into the liquor between the porous cathode cells and the 
metal baskets. 

The production of one kg. of permanganate requires 
about 7 K.W.H. Voltage is about 2'S volts, cathodic I.D. 
is '85 amp. per cm 2 , and anode I.D. is '085 amp. per cm 2 . 
Nickel electrodes are found very satisfactory. 1 

Potassium Ferricyanide* This substance which was 

FIG. 57. 

formerly produced by the action of chlorine upon potassium 
ferrocyanide is now generally made by electrolytic oxidation 

2K 4 FeCy 6 + O + H 2 O = 2K 3 FeCy 6 + 2KOH. 

A saturated solution of the ferrocyanide is used, made 
slightly alkaline, at a temperature of 20 C. 

Nitric Acid by Electrolysis of Peat Deposits? The process 
invented by Nodon seems to have been adopted in a few 
districts where peat deposits exist. It depends upon the fact 
that in such deposits there is a considerable quantity of 
calcium nitrate dissolved in the water. Carbon anodes and 
iron cathodes are sunk into the deposit, connected up in sets, 
and drainage arranged to convey the nitric acid, produced at 

1 Zeitsch. Elektrochem., 1910, 16, 170. 

2 Zeitsch. anorg. Chem., 1904, 39, 240. Eng. Pat., 7426 (1886). 
Electrical Review, 1893, 32, 216. 

3 Met. and Chem. Eng., 1914, 12, 107. 


the anodes, to storage tanks. The nitrifying bacteria are not 
destroyed by the process, and continue their work, uninter- 
ruptedly, of converting the nitrogenous matter in the peat 
into nitrates. 

Sewage Disposal^ Electrolysis has been applied to 

FIG. 58. 

sewage disposal in the Landreth process employed at New 
York. Treatment with lime, and a subsequent electrolysis 
in special towers or partitioned tanks render the fluid 

Fluorine. This element was isolated by Moissan in 1887 

1 Met. and Chem. Eng., 1915, 13, 735, 993. 


by electrolysing anhydrous hydrofluoric acid containing potas- 
sium fluoride, in an apparatus made entirely of platinum. 
He subsequently proved that copper may replace platinum 
as a containing vessel since it becomes coated with copper 
fluoride which protects the metal from further corrosion 
Apparatus was designed by Moissan for producing fluorine 
on the large scale, and the plant is made by MM. Poulenc 
Freres of Paris. 1 The hydrofluoric acid is contained in a 
copper vessel B (Fig. 58), the inner surface of which, C, acts 
as cathode. B is inside a larger vessel S, which contains a 
cooling mixture, and it is covered by a copper lid M, from 
which it is insulated by a rubber ring L, and also by rubber 
tubes round the bolts b. A is a copper tube perforated below 
by a number of small holes d. This serves as a diaphragm 
between the cathode and the anode /, which is of platinum 
and surrounds the copper tube T, which communicates with 
the upper vessel N containing a cooling mixture. The 
bottom of the anode chamber is covered by a copper plate 
g which is fastened to T by copper screws v. The diaphragm 
which is in electrical connection with the anode becomes 
at first covered with a layer of copper fluoride, and then 
being insulated, fluorine is evolved from the anode. Copper 
tubes H, F, serve to carry a.way the evolved hydrogen and 

1 Zeitsch. Elektrochem., 1900, 7, 150. 



MANY organic compounds may be produced by electro- 
lysis, but since this branch of manufacturing chemistry 
has had its chief home in Germany, it is difficult to find 
out to what extent electrolytic methods have actually been 
applied. During the twenty-five years ending 1910, for 
example, 1 about 100 patents were taken out for the pre- 
paration of organic products by electrolysis. Of these, 89 
were German, 6 were French, 3 English and 4 American 
and 39 of the German patents referred to the electrolytic 
reduction of organic compounds. This branch of electro- 
chemistry has not received, the attention which it deserves 
in countries other than Germany. In these processes there 
are many variables to be taken into account, such as, 
current density, electrode material, overvoltage, etc., which 
indicate complexity, but the complexity foreshadows great 

A brief account will now be given of processes well estab- 
lished on a large scale, and also a summary of original work 
and patent literature which will indicate the direction that 
progress in this branch has taken. 

lodoform. The older method of producing this important 
antiseptic, by the action of iodine upon alcohol in the presence 
of sodium carbonate, was based upon the following reaction 

CH 3 .CH 2 OH + 8Na 2 C0 3 + 5!, + 2H 2 O = CHI 3 

+ 9NaHCO 8 + /Nal. 
Only about 30 per cent, of the iodine goes to form iodoform, 

1 Met, and Chem. Eng., 1915, 13, 211. 


but by electrolysing alkali iodide in presence of alcohol the 
following reaction takes place 

CH 3 .CH 2 OH + sNal + 3H 2 O = Na 2 CO 3 + CHI 3 

and all the iodine used goes into the iodoform. 1 The con- 
ditions for preparing iodoform in this way are as follows : 2 
Sodium carbonate 50 parts and potassium iodide 170 parts, 
are dissolved in 96 per cent, alcohol, 100 parts. The anode 
should be of smooth platinum, and the cathode of lead is 
encased in parchment or some suitable diaphragm material. 
Working temperature should be 60-70 C., anode I.D. 1-2 
amps, per dm 2 ., and voltage 2-2-5 volts. Iodoform is pro- 
duced at the rate of 1*3 gms. per amp.-hour, or 500 gms. per 
K.W.H., and the current efficiency is about 90 per cent. A 
satisfactory yield of iodoform is also obtained if acetone be 
used in place of alcohol. 

Bromoform can be obtained in a similar manner if alkali 
bromide be electrolysed in the presence of alcohol or 
acetone. 3 

Anthraquinone. The first step in the direction of apply- 
ing electrolytic methods to the oxidation of anthracene was 
made when the " spent " chromic acid liquor was revived by 
oxidation, in the anode compartment of a two compartment 
cell, being then returned to the oxidising vessel for converting 
a fresh quantity of anthracene to the quinone. The condi- 
tions laid down by the Farbwerke Hochst Patent 4 were : 
Solution to contain about 100 gms. of Cr 2 O 3 per litre with 
350 gms. of sulphuric acid, current density at anode of lead 
about 3 amps. The anode functions as a PbO 2 electrode 
which Miiller and Soller have shown acts catalytically. 5 

The reaction taking place is 

Cr 2 (S0 4 ) 3 + 30 4- 5H 2 O = 2H 2 CrO 4 + 3H 2 SO 4 , 
and a current efficiency of about 80 per cent, is obtained. 

1 D.R.P., 29771 (1884); Eng. Pat., 8148 (1884). 

2 Zeitsch. Elektrochem., 1897, 3 > 268 - 

8 Atner. Chem.Journ., 1902, 27, 63; Zeitsch. Elektrochem.^ 1904, 10, 
409; Trans. Amer. Electrochem., 1905, 8, 281. 

4 D.R.P., 103860 (1899). Zeitsch. Elektrochem., 1900, 6, 290, 308. 
6 Zeitsch. Elektrochem.^ 1905, 11, 863; 1913, 19, 344. 


Askenasy has shown that the diaphragm may be dispensed 
with if certain salts are added to the solution, such as sulphate 
or oxalate of sodium, or magnesium sulphate. 

In the process used by Farbwerke vorm. Meister Lucius 
und Briining, 1 chromic acid is not used, but instead, 20 per 
cent, sulphuric acid electrolyte containing 2 per cent, of cerium 
sulphate. The temperature is kept at 70-1 00 C, a lead-lined 
vessel forms the anode and the liquor is well stirred. 

Anode I.D. is 5 amps., at a pressure of 2'8 to 3*5 volts, 
and current efficiency is nearly 100 per cent. 

Vanillin. This substance is produced by electrolytic 
oxidation of the sodium salt of isoeugenol, by v. Heyden 
Nchfg. 2 in Germany. A 15 per cent, solution of the sodium 
salt in excess of soda fills the anode cell, and the cathode 
chamber contains 1020 per cent, caustic soda ; working tem- 
perature 60 C. The anodes of PbO 2 act as catalysts, since 
platinum anodes merely evolve most of the oxygen without 
changing the substance. 3 

/OH /OH 

C 6 H 3 -OCH 3 + 30 = C 6 H 3 -OCH 3 + CH 3 COOH. 
\CH : CHCH 3 \CHO 

Isopropyl Alcohol. This is produced by the reduction of 
acetone. The claims of Merck's patent have been verified by 
Elbs. 4 The cell is divided into two compartments, and with 
a mercury cathode the yield of alcohol is good, very little 
pinacone being formed. 

Chloral is produced when alcohol is allowed to drop into 
the anode compartment of a cell in which potassium chloride 
is electrolysed. 5 

Saccharine is produced by the electrolytic oxidation of 
0-toluene sulphonamide 6 

p R /SO 2 NH 2 , ~ _ PR / ^^2\MR _i_ R o 
^ n 4 - ^i*H 2H 2 u. 

1 Electrochem. Ind., 1904, 2, 249. 

2 D.R.P., 92007 (1895). Electrochem. Review, 1900, 1, 31. 

3 Electrochem. Ind., 1904, 2, 452. Austrian Pat., 34562 (1908). 

4 D.R.P., 113719 (1899). Zeitsch. Elektrochem., 1902, 8, 783. 
8 Elektrochem. Zeitsch., 1894, 1, 70. 

D.R.P., 85491 (1895). 


Sugar juices can be purified by electrolysis. 1 
The processes thus enumerated are interesting and im- 
portant applications of electrolysis to the production of 
organic compounds. A full account of the electro-chemistry 
of organic compounds will be found in the treatise by Dr. 
Lob 2 on the subject. 

The following summary indicates broadly the main pro- 
vinces which have been explored experimentally, and, judging 
by the patent literature, this pioneering work has been utilised 


The numerous reduction products of nitro-benzene and its 
homologues can be obtained by electrolytic reduction. Many 
patents have been issued specifying the use of particular 
cathode metals. 

Elbs has proved that with attackable cathodes reduction 
proceeds further than with unattackable cathodes, such as 
platinum nickel or mercury. 

The catalytic action of various salts in solution has also 
been proved of great use in aiding or directing cathodic 

Oxidation, although it has received much attention, has 
not the same degree of importance as reduction. The side 
chains of aromatic compounds can be oxidised in such a 
manner as to give the groups -CH 2 OH, -CHO, -COOH. 

Many dye-stuffs have been prepared by electrolytic 
reduction or oxidation. A few examples are 

/-Rosaniline, by reduction of /-nitro-diamidotriphenyl- 
methane in concentrated acid, D.R.P. 84607 (1894). 

Orange dyes by using as cathode liquor an alkaline 
solution of the yellow condensation product of /-nitrotoluene 
sulphonic acid, Eng. Pat., 22482 (1895). 

The electrolytic diazotisation of amines and preparation 
of azo-dyes was discovered by Lob. 3 The amine, nitrite, and 

1 Jahrbuch der Elektrochemie, 1901, 8, 628. 

2 The Electro-chemistry of Organic Compounds (1905). 

3 Zeitsch. Elektrochem., 1904, 10, 237. 


coupling compound, in neutral or alkaline solution* form the 
anode liquor, and the mixture is electrolysed, using a platinum 
or some other unattackable anode. The anodic mixture 
should be stirred during electrolysis and cooling is not neces- 
sary, since, at the moment of diazotisation by the discharged 
nitrite ion, the diazo-compound condenses with the phenolic 
coupling material. For example, by using an anode liquor 
containing the sodium salt of sulphanilic acid, nitrite and 
/ft-naphthol, the dye, Orange II, is obtained, similarly dian- 
isidine blue is formed from dianisidine, nitrite and /?-naphthol. 

The electrolytic oxidation of amines results in the produc- 
tion of colouring matters, and many have been prepared in 
this way. Aniline black is obtained from aniline by this 
method ; also naphthylamine violet from the base. Mixtures 
of the bases also have been oxidised with similar results. 1 

The reduction of the carbonyl group to (CHOH) is well 
illustrated by the formation of borneol from camphor and 
also benzhydrol from benzophenone. 

Substitution by the halogen elements also is very general 
as shown by the formation of chloral, chloroform, bromo- 
form, chloraniline and the electrolytic method of carrying out 
Sandmeyer's reaction. 2 

The application of electrolysis to the production of organic 
compounds awaits development in many directions and offers 
a rich field for investigation. 

1 Zeitsch. angew. Chemie, 1894, p. 107. 

2 Zeitsch. Elektrochem., 1901, 7, 877. 


Acker cell for caustic soda, 112 
Alternating current, transmission 
of, 27 

use of, in electrolysis, 52 

Alumina, purification of, 55 

Aluminium, production of, 53 

Ammeter, 28 

Anodes, 13 

Anthraquinone, production of, 140 

Antimony, production of, 77 

refining of, 45 

Ashcroft process for sodium, 72 

Balbach-Thum silver process, 48 
Bayer process forpurifyingalumina, 


Becker process for sodium, 71 
Bell process for caustic soda, no 
Belt's lead-refining process, 42 
Billiter-Leykam cell for caustic 

soda, in 
Billiter-Siemens cell for caustic 

soda, 102 

Bismuth, production of, 78 
Borcher's cell for electrolysis of 

fused salt, 72 
Bromine and bromates, production 

of, 127 

Bromoform, 140 
Browne process for nickel, 65 
Brunner, Mond & Co., process for 

zinc, 61 

Cadmium, refining of, 45 
Calcium, production of, 75 
Calorie, electrical equivalent of, 7 
Castner cell for caustic soda, 105 

process for sodium, 66 

Castner - Kellner cell for caustic 

soda, 1 06 
Cataphoresis, 9 
Cathodes, 13 
Caustic soda from brine, 91 

from fused salt, 112 

Chloral, preparation of, 141 

Chlorates, production of, 116, 123 
Chlorine and caustic soda, 91 
Chrome yellow, preparation of, 131 
Chromic acid, production of, 141 
Colloids in electrolysis, 10 
Commutator, 24 
Copper, analysis of crude, 34 

production of, 57 

refining of, 33 

Coulomb, 6 

Cryolite, density of fused, 55 
Cuprous oxides, production of, 131 
Current efficiency, n 

measurement, 28 

Cyanide gold-refining process, 52 

DanielPs cell, 17 

Darling process for sodium, 69 

Decomposition voltage, 7, 8, 80 

Diaphragm cells for caustic soda, 92 

Diaphragms, 13 

Dietzel silver refining process, 49 

Dow cell for bromine, 128 

Dry cells, 1 8 

Dyestuffs, production of, 142 

Dynamo, construction of, 23 

Edser-Wilderman cell for caustic 

soda, 109 
Electrical osmosis, 9 

units of measurement, 6 

Electrodes, 12 
Electrolysis, I 

bath, 12 

Electromotive force, 6 
Energy efficiency, n 

Faraday's laws, 3 

Filter-press cells for electrolysis of 

water, 82, 88 

Finlay cell for caustic soda, 99 
Fluorine, production of, 137 
Fluosilicate lead-refining process, 


Foul electrolytes, analysis of, 40 
Fuel cells, 18 




Garuti process for hydrogen and 
oxygen, 85 

Gibb's process for chlorate produc- 
tion, 124 

Gold refining, 50, 52 

Griesheim Elektron cell, 92 

Haas-Oettel cell for hypochlorite, 


Hargreaves-Bird cell for sodium 

carbonate, 94 
Heat of formation, 7, 79 
Hermite process for hypochlorite, 


Hoepfner process for copper, 59 

process for nickel, 64 

process for zinc, 61 

Hydrogen and oxygen, production 

of, 79 

peroxide, production of, 134 

Hydrosulphite, production of, 133 
Hydroxylamine, production of, 132 
Hypochlorites, production of, 116 

International oxygen cell, 87 
Iodine, production of, 129 
lodoform, production of, 139 
Ions, 2 

Iron refining, 45 

Isopropyl alcohol, production of, 

Joule, value of the, 7 
Joule's law, 3 

Keith's process for lead refining, 42 
Kellner air-pressure cell, 107 
Kellner cell for hypochlorite, 119 
Kilowatt, value of the, 7 
Kossuth cell for bromine, 128 

Landreth process for sewage dis- 
posal, 137 

Laszcynski process for zinc, 61 
Lead, electrolytic production of, 63 
peroxide, 131 

refining, 41 

Leclanche cell, 17 

Lelande cell, 18 

Le Seur cell for caustic soda, 101 

Lithium, production of, 77 

Load factor, 31 

MacDonald cell for caustic soda, 

Magnesium, production of, 74 

Mansfeld process for production of 

copper, 58 
Marchese process for production of 

copper, 57 

Mercury cells for caustic soda, 103 
Metal fog, 14, 56 

Moebius silver-refining process, 46 
Molten electrolytes, use of, 14 
Motor generator, 27 

Nickel, production of, 64 
Nitric acid from peat, 136 

Ohm, the unit of resistance, 6 
Ohm's law, 6 
Osmotic pressure, 4 
Outhenin-Chalandre cell for caustic 

soda, 95 
Overvoltage, 9 
Oxidation of organic compounds, 

Ozone, production of, 89 

Peat, dehydration of, 9 
Percarbonates, production of, 132 
Perchlorate, production of, 116, 

process for lead refining, 44 

Permanganate, production of, 135 
Persulphuric acid, production of, 


Polarisation, 17 
Potassium ferricyanide, 136 
Potential, 6 
Power costs, 30 

sources, 29 

transmission, 27 

Primary cells, 16 

Reduction of organic compounds, 


Rhodin cell for caustic soda, 108 
Rotary converter, 26 

Saccharine, production of, 141 

Salom's lead process, 63 

Schmidt process for hydrogen and 

oxygen, 82 
Schoop's process for hydrogen and 

oxygen, 84 

process for hypochlorite, 122 

Schuckert cell for hypochlorite, 121 
process for hydrogen and 

oxygen, 87 
Secondary cells, 19 



Sewage purification by electrolysis, 

Siemens and Halske process for 

copper, 58 

Silver, refining of, 46 
Slimes from copper refining, 39 
Sodium chloride, electrolysis of, 

from fused caustic soda, 66 

from fused salt, 72 

from fused sodium nitrate, 69 Water, electrolysis of, 79 

contact electrode process for, power, 31 

Watt, electrical value of the, 6 
Weston cadmium cell, 6 
White lead, production of, 130 
Whiting cell for caustic soda, 107 
Wohlwill process of gold refining 

Townsend cell for caustic soda, 96 
Transformers, 27 

Units, electrical, 6 

Vanillin, preparation of, 141 
Vautin cell for caustic soda, 114 
Voltaic cell, 16 
Voltmeter, 28 

Solvay-Kellner cell for caustic soda, 

Sugar juice, electrolytic purification 

of, 142 
Swinburne and Ashcroft process for 

zinc chloride, 62 

Thermopiles, 20 
Tin, refining of, 44 
Tommasi process for lead refining, 

Zinc chloride, dehydration of, 62 
-- , electrolysis of fused, 62 

- electrolytic production of, 60 

- white, electrolytic production 
of, 131 


ACKER, 112 
Addicks, 45 
Allmand, 99 
American Smelting and Refining 

Co., 45 

Anaconda Copper Co., 38, 61 
Archibald & Finlay, 99 
Arrhenius, 5 
Ashcroft, 31, 62, 72 

Baekeland, 98 
Balbach Works, 48 
Bancroft, 41 
Betts, 10, 42, 78, 126 
Bolton & Sons, 33 
Borchers, 42, 75, 78 
British Aluminium Co., 54 

Thomson-Houston Co., 2 

Brunner, Mond & Co., 61 
Burgess, 130 

Canadian Copper Co., 65 

Smelting Works, 43 

Carrier, 71, 74 
Castner, 66, 104 

Alkali Co., 66, 108 

-Kellner Co., 63, 66, 104 

Clevenger, 52 

Consortium f. Elektrochemische 

Industrie, 134 
Constam, 132 
Couleru, 127 

Darling, 69 
D'Arsonval, 81 
Davy, i 

Del Proposto, 85 
Deville, 53 
Dietzel, 49 
Donnan, 31 

Edser-Wildermann, 109 
Elbs, 134, 141, 142 

Electrical Lead Reduction Co., 64 
Electrolytic Alkali Co., 94 
Elkington, 23, 33 
Elliot's Metal Co., 33 

Faraday, I, 2, 23 

Farbwerke vorm. Meister, Lucius 

& Briining, 141 
Foerster, 78, 123, 129 
Frary, 77 

Garuti, 85 

Gibbs, 124 

Goodwin, 77 

Gramme, 23 

Great Falls Refinery, 40 

Griesheim Elektron Co., 71 

Griinauer, 63 

Haber, 56, 115 
Hall, 53 

Hambuechen, 130 
Hermite, 120 
Heroult, 53 
Hevesy, 69 
Hoffvan't, 4 

Imhoff, 124 

International Oxygen Co., 87 

Jellinek, 133 

Keith, 42 
Kellner, 107, 119 
Kershaw, 31, 32, i 2 JI 5 

Laszcynski, 57, 6 1 
Latchinoff, 81 
Leblanc, 71 
Lederlin, 125 
Le Seur, 101 
Lob, 142 


Lorenz, 15, 62, 75 
Luckow, 130 

Macdonald, 101 
Mansfeld Copper Co., 58 
Marchese, 57 
Moebius, 46 
Moissan, 137 
Morse, 4 
Mott, 77 
Moynot, 115 
Mueller, n, 123 

Neumann, 56 

New York and Pennsylvania Co. 

Paper Mills, 101 
Niagara Alkali Co., 102 
Nichol's Refinery, 35 
Nodon, 136 
Norddeutsche Affinerie, 38, 50 

Oettel 74, 121 

Olson, 56 

Orford Copper Co., 66 

Parker & Robinson, 129 

Pascal, 56 

Patten, 77 

Pfeffer, 4 

Philadelphia Mint, 47, 51 

Pietzsch & Adolf, 134 

Plato, 75 

Poplar Borough Council, 122 

Poulenc Freres, 138 

Pring, 32 

Pyne, 55 

Raritan Copper Works, 48 
Rhodin, 108 
Richards, 52, 74, u'5 
Richardson, ^6 
Rinck, 129 
Ruff, 75, 77 

Sadtler, 63 

Salom, 63 

Salzbergwerke neu Stassfurt, 135 

Savelsburg & Wannschaff, 65 

Schering, 135 

Schmidt, 82 

Schoop, 84, 122 

Schuckert, 87, 121 

Seward & von Kiigelgen, 73, 76 

Siemens & Halske, 52, 58, 77, 102, 

Siemens, Schuckert Co., 121 

Spear, 10 

Standard and Colorado City- 
Works, 1 01 

Steinhardt & Vogel, 62 

Swan, 32 

Swinburne, 62 

Tafel, 132 
Taussig, 95 
Thompson, 56 
Threlfall & Wilson, 126 
Tommasi, 42 
Townsend, 96, 115 
Tronson, 77 
Tucker, 77 
Tuttle, 51 

Ulke, 41 

United Alkali Co., 62, 73 

Volta, 1 6 

Walker, 52, 129 
Watt, 33, 9i 
Whiting, 107 
Whitney, 77 
Winteler, 126 
Wohler, 76 
Wohlwill, 50 
Woolrich, 23 






WAV 1 


LD 21-50m-l,'33